Imagine a hunting animal whose visual acuity is similar to a cat’s, even though it’s more than 100 times smaller and equipped with a brain with only half the neurons of a cockroach. If you watch it stalk its prey, it’s cat-like too – crouching and approaching slowly before pouncing from a distance, jumping about five times its own body length.
The main difference is that this animal is doing it on more legs, says Ximena Nelson, a biologist at Canterbury University.
Yes, it’s a jumping spider. One of the jumping spider’s other endearing behaviours is that if you look at it, it will turn its head and look straight back at you.
But it is the spider’s hunting behaviour that is of most interest to biologists studying cognition. As it moves around its prey, it sometimes even loses sight of it before attacking. This suggests jumping spiders have some sort of memory or spatial ability to know where their prey is in relation to them and that they assess their options, comparing the rewards and risks, and form an attack plan.
The question Nelson wants to answer is whether jumping spiders are able to plan ahead like much larger-brained mammals. She recently became one of very few New Zealanders to receive funding from the international Human Frontier Science Program to investigate the predatory behaviour of Portia jumping spiders, which live in tropical forests and feed on other spiders, and whether good vision or their habitat determines their planning.
Unlike insects, spiders don’t have compound eyes. Of all invertebrate species, it’s only cephalopods (octopuses, squid) and jumping spiders that have camera eyes similar to ours.
Excellent vision has allowed jumping spiders to forgo web spinning. “They’ve essentially emancipated themselves from webs because their vision enables them to exploit their habitat much more fully than a web does,” says Nelson.
Portia spiders are an exception but it’s only the female that makes a temporary “scrappy web” to lay her eggs.
The first challenge for Nelson will be to design an experimental route for the spiders that restricts it from backtracking on bad decisions. In experiments, she’s already built mazes to test their forward thinking – but the spider outsmarted her by taking shortcuts.
We already know from bees and ants that social insects have evolved phenomenal behaviours, including the waggle dance bees use to communicate the location of a good food source to their hive mates. But such behaviours are optimised by the social collective, whereas “jumping spiders are famously antisocial”, Nelson says.
Both bees and ants are also really good at spatial navigation, finding their way back home from long distances. But for jumping spiders, it’s more difficult to determine what clues they use.
“This research is about trying to understand the factors that seem to select for, or lead to, jumping spiders having this ability to plan ahead,” she says.
Jumping spiders live just about everywhere, from 7000m up Mt Everest to beneath sand dunes. One of Nelson’s hypotheses is that their habitat needs to be of medium complexity, allowing them to both see across long distances and hide. In these habitats, planning may be possible, not only in mammals and birds with large brains, but also in small animals with tiny brains, she says.
“The whole point of doing this with Portia spiders is because here we’re talking about an animal that’s got a small fraction of the neurons of a mammal and yet it’s displaying mammal-like behaviour. It seems to be doing all of this computational stuff with far less computing elements.”
The findings may be significant because they could lead to the development of machine learning algorithms for artificial planning systems in machines with power constraints, such as rovers used in space missions, so they can do more with less.