Fish robots put a scare into invasive species

Invasive mosquitofish (Gambusia holbrooki) have been wreaking havoc in Australian freshwater systems since their arrival from the US in the 1920s. Introduced as a means of natural mosquito control, the fish turned out to be no better at this job than their native counterparts. And following the narrative of most biological introductions (cane toads are a prime example), the fish brought with them a suite of problematic behaviours that have significantly damaged ecosystems, and that researchers have been struggling to address ever since.

Now, an interdisciplinary team of biologists and engineers from Australia, the US and Italy has hit upon a promising new method of controlling populations of these pests – a terrifying robot fish that scares them senseless.

Despite sounding like the backstory to a horror movie, the team is working on solid science. Publishing in the journal iScience, the researchers outlined how they designed their robot to closely mimic the appearance and movements of the mosquitofish’s main predator, the largemouth bass (Micropterus salmoides).

‘We made their worst nightmare become real: a robot that scares the mosquitofish but not the other animals around it,’ says first author Giovanni Polverino of the University of Western Australia.

Aided by computer vision, the robot struck like a real predator when it spotted the mosquitofish approaching tadpoles of the motorbike frog (Litoria moorei), a species threatened by mosquitofish. This constant perceived threat of predation resulted in marked behaviour changes in mosquitofish. In aquaria trials, mosquitofish tended to huddle together at the centre of the testing arena, hesitant to venture into exposed and uncharted waters. They also swam more frenetically, with frequent and sharp turns, than those that hadn’t encountered the robot.

The fearfulness seems to last. When shifted back to safe, robot-free aquaria, the shaken-up mosquitofish showed signs of anxiety that lasted for weeks after their close robot encounter. Scared and stressed, the mosquitofish were less active, ate more, and froze for longer periods, checking for danger.

After five weeks of brief encounters between the fish and the robot, the researchers found that the entire physiology of the mosquitofish had shifted, with energy distributed away from reproduction and towards basic survival. The bodies of males became thin and streamlined with stronger muscles near the tail, built for rapid escape. Male fish also had lower sperm counts, while females produced lighter eggs, which are changes that are likely to compromise the species’ survival in the long term.

But what about the tadpoles that the mosquitofish usually prey on? Were they terrified by this marauding robot, too?

Luckily, tadpoles have very poor eyesight: they could barely see the robot. This didn’t mean they were unaffected, though.

‘We expected the robot to have neutral effects on the tadpoles, but that wasn’t the case,’ says Polverino.

With the mosquitofish too scared to venture out to nibble their tails, the tadpoles were free to venture out into the testing arena.

‘It turned out to be a positive thing for tadpoles,’ says Polverino. ‘Once freed from the danger of having mosquitofish around, they were not scared anymore. They’re happy.’

So why don’t we simply use real largemouth bass? History tells us that deliberately introducing new species as a means of biological control is fraught with issues – don’t forget, that’s how the mosquitofish got here in the first place. A recent analysis has estimated that introduced invasive species have cost the Australian economy $389.59 billion in the last 60 years.

So clearly, introducing further non-native species in the hope that they will help us to reign in our current invasive species problems is asking for trouble. But current methods aren’t adequately addressing the problem, either.

‘Mosquitofish is one of the 100 world’s worst invasive species, and current methods to eradicate it are too expensive and time-consuming to effectively contrast its spread,’ says Polverino. “Instead of killing them one by one, we’re presenting an approach that can inform better strategies to control this global pest.’

It will be a while yet before you might spy a robot fish at your local freshwater fishing hole. The team still needs to overcome a few technical challenges before progressing beyond the aquaria trials.

‘While successful at thwarting mosquitofish, the lab-grown robotic fish is not ready to be released into the wild,’ says senior author Maurizio Porfiri of New York University. As a first step, the researchers plan to test the method in small, clear pools in Australia, where two endangered fish are threatened by mosquitofish.

‘Invasive species are a huge problem worldwide and are the second-highest cause for the loss of biodiversity,’ says Polverino. ‘Hopefully, our approach of using robotics to reveal the weaknesses of an incredibly successful pest will open the door to improve our biocontrol practices and combat invasive species. We are very excited about this.’