NPR logo Putting Spiders On Treadmills In Virtual-Reality Worlds


Putting Spiders On Treadmills In Virtual-Reality Worlds

Jumping spider.

The methodology described in a recent study, a peer-reviewed paper on animal behavior — in which biologists Tina Peckmezian and Phillip W. Taylor of Macquarie University in Sydney, Australia, worked with 56 female jumping spiders — is fascinating.

Here's what the biologists did:

First, the spiders were transferred from their cages to a refrigerator where they were "cooled until quiescent," as Peckmezian and Taylor put it, and placed on a chilled granite block. Then, small magnets were "gently affixed" to the spiders using a bit of dental cement, "taking care not to cover the eyes."

After a 24-hour period of recovery, the spiders were then "gently" lifted, using special equipment, onto a 3D-printed spherical treadmill that was positioned in front of a display screen, where the spiders were held in place by their magnets.

Next, the biologists projected a virtual-reality (VR) environment onto the screen that the spiders could walk through. (Not quite The Matrix or Plato's "Allegory of the Cave" that Marcelo wrote about last month, but an arachnid equivalent.)

And, here's the best part of all: It was a closed loop VR, meaning a VR environment in which the spiders' own movements caused contingent changes in what happened on the screen. (Previous experiments with jumping spiders occurred using an open loop system in which stimuli are presented independent of the animals' responses, which decreases the "reality" aspect of virtual reality.) The closed-loop projected scenes were set against a flat ground plane that was in texture and color meant to resemble tree bark.

Article continues after sponsorship

Now, if you're like me, you have to be asking: "Why?"

Why put spiders on little treadmills in VR worlds?

The immediate goal of this research was to see if the spiders' behavioral tendencies and learning transferred from RW (the real world) to VR. The result was clear: They did. In one experiment, spiders encountered either an empty virtual world or one with 18 pillars randomly placed within it. Spiders in the empty condition were significantly less active than those in the more complex environment. The spiders' individual tendencies for spontaneous locomotion and other behavioral preferences were conserved in VR compared to RW.

In another experiment, spiders in RW encountered one of two conditions: either a beacon (a red pillar or green cross) placed behind their nest site or no beacon. Once put in VR, the spiders with beacon experience made more visits to the beacons — positioned in the same place relative to their own orientation as had been the case in RW — and spent more time near the beacons than spiders who hadn't first learned about beacons in RW. A learned association between beacons and nest sites had, in other words, transferred successfully across the two types of environments, just like spiders' individual tendencies had done.

These results are fun to know about in themselves, but important primarily because they validate that the VR environment will be a useful tool for the researchers' long-term goals: Peckmezian and Taylor seek new ways of studying visually mediated cognition in invertebrates.

As it turns out, jumping spiders are the perfect candidate for this endeavor because, unlike insects with their compound eyes, they have four pairs of specialized eyes, each with a single lens. Jumping spiders see the world with depth perception, color vision and, as the researchers write in their article, "a retina with spatial acuity that greatly exceeds that of any other animal with eyes of comparable size."

Earlier this week, I asked Tina Peckmezian by email if she could expand a little bit on the significance of their research plans. She responded in part:

"I can appreciate that the idea of spiders wandering around through virtual space may sound like science fiction, but virtual reality technology has been an incredible tool in the behavioral and neurosciences. A common goal for animal behavior researchers is to find a middle ground between providing realistic, ecologically relevant experimental conditions [and] conditions that can be precisely controlled and reproduced across trials.

Typically, there is a trade-off; you can test animals in their natural environment but can't control the environmental variables present on a particular day; or you can present a series of automated stimuli, but these don't provide important feedback information to the responding animal. VR lets you do both. You can create immersive, highly realistic 3D virtual environments that update in real time as the animal moves through them, and, since animals are typically restrained, can couple these with neurophysiological recordings."

So, now we know: Jumping spiders navigate VR worlds with arachnid aplomb.

Even as you read these words, more jumping spiders may be chilling in refrigerators, ready to take their places on the little treadmills. As they do, scientists will learn more about invertebrate perception and neurophysiology.

Barbara J. King, an anthropology professor at the College of William and Mary, often writes about human evolution, primate behavior and the cognition and emotion of animals. Barbara's most recent book on animals is titled How Animals Grieve. You can keep up with what she is thinking on Twitter: @bjkingape