Driving fish?

Julia Ravey speaks with Ohad Ben-Shahar and Ronen Segev of Ben-Gurion University of the Negev who have been studying the cognitive behaviour of our famously forgetful sea-dwelling friends...

Julia - Call me strange, but I find driving a car to be really relaxing. It gives me a focus with a bit of peace and quiet. But, in the future, maybe we won't be driving anymore. Various tech giants are beavering away trying to perfect automated vehicles, but what they maybe haven't considered in their grand plans is goldfish. I was scrolling on Twitter and one of the most bizarre yet fascinating videos popped up in my feed of a goldfish steering a tank on wheels with surprisingly good accuracy, and this was all in the name of science. I reached out to Ohad Ben-Shahar, a researcher behind this work, to see if carp-ools could be what the future of driving has in store.

Ohad - Julia, you might find it shocking, but fish don't drive cars. Trying to make them do so, even though it sounds cool, was not our end goal. In fact, making fish drive your robotic vehicle is part of our broader agenda to explore how animals represent the surrounding space in their brain and how they make navigational decisions.

Julia - Ohad and fellow researcher, Ronen Segev, wanted to get the goldfish on the road to test their navigational skills. But, as Ronen explains, there is one problem with wanting to teach a fish to drive.

Ronen - Okay, clearly you cannot ask a fish to do that. So you need to train it in stepwise manner.

Julia - The researchers used their powers of persuasion to get the fish behind the wheel: a target and some fish food. The goldfish were placed in a special robotic tank on wheels, which moves according to where the fish swims. When the fish tank hit the target on the wall, the goldfish would be given a treat, teaching them that this target was one to chase.

Ronen - After training, which is usually 10-15 sessions, each one of them 30 minutes in total, they actually managed to go directly towards the target in order to consume as many food pellets as they could during the 30 minute session.

Julia - And it seems the goldfish were specifically eyeing up the target: they managed to seek it out even when the researchers were being a little bit fishy.

Ronen - One thing to do is to take the target and place it on the other side of the room. So, initially, what you see is that the fish goes to the previous location, but then it actually finds out that the target has moved and it goes directly to the new target. So, this thing that we do in the experiment, actually manipulating the arena where the fish needs to navigate, actually makes us believe, and it's very convincing, that the fish indeed controls the vehicle, managed to drive the vehicle, and also remembered the surroundings and navigated towards the target.

Julia - Do we think fish use this type of navigation - remembering landmarks or cues - in the wild.

Ohad - Visual landmarks must play a role in their navigational decisions just as it helps us humans. But I want to be slightly careful here: we have currently studied only goldfish. So, at the moment, we can't really generalise to all species of fish, but studying other species is indeed part of our plan in the near future.

Julia - Ronen and Ohad want to go further than just looking at the goldfish's movements in this somewhat bizarre driving scenario, and also peek into their brains when they navigate in open waters.

Ohad - What we ourselves are currently developing is the ability to record the activity of single cells in the fish's brain: Why it is behaving completely naturally in its environment? We hope to find the links, if you will, between behavior and the neural coding, and hopefully understand how navigation decisions are indeed represented and made in the fish's brain.

Julia - The results from this experiment, to me, makes it seem like goldfish are pretty sophisticated if they can do this type of navigation. Should we all stop using the phrase, "You have a memory like a goldfish", when someone's a little bit forgetful.

Ronen - Yes, of course. As someone who has worked on fish for many years, these animals are amazing! They constitute something like 35,000 different species and they do amazing things. So, definitely, we should stop thinking that fish have low cognitive capabilities or things like that. They just live in a very different world than us and they solve different things differently.

Julia - But we've got no chance of water-mated vehicles coming along anytime soon?

Ronen - No, I don't think so.