The Lion(fish)'s Share: Robotic Aid Against Invasive Red Lionfish

In the warm waters off the coast of Florida, Bermuda, and the Carribean, a fascinating new member of the food chain is making quite a splash. Though instead of sporting the expected scales and fins, it's made mostly of plastic tubes and wiring. Being hailed as the 'Roomba of the Sea', these robotic constructs have a prime objective: ridding the waters of invasive lionfish.

A lionfish: Image Credit: Tim Proffitt-White
https://www.flickr.com/photos/tim_proffitt_white/

Lionfish (which includes 12 species under the genus Pterois) are beautiful creatures, and have become something of a hallmark ocean species. While not known for their taste, they are popular inhabitants of saltwater aquariums around the world. The different species of lionfish vary in distribution, size, coloration, and so forth, and most are considered model citizens of the ecosystems they inhabit. However, when they escape their native ranges into unsuspecting waters, they can quickly overwhelm their new surroundings. One of the more troublesome species of lionfish, known as the red lionfish (Pterois volitans,) has made a name for itself as a destroyer of reefs.

Traditionally restricted to the Pacific and Indian Oceans, lionfish are fairly common from South Africa to Australia to the southern coasts of Japan, and everywhere in between. They are characterized by bands of red, white, cream, or black stripes (which signal their venomous nature,) as well as their long intricate fins which kind of, I suppose, resembles a lion's mane. A series of venomous fin rays, most located along the spine, can deliver a dose of venom which deters any potential predators, and can have harmful effects on any unfortunate humans who end up in their path (though luckily their sting is rarely fatal.)

The big problem with lionfish in the Atlantic is, unfortunately, their mere existence. With a voracious appetite, and quick reproduction habits, lionfish reduce the native fish populations of reefs drastically. So much so that algaes, which are normally kept at bay by the aforementioned native fish population, begin to take over reef systems and choke out the coral which build them.

Crossing Over (Without John Edward)

How did lionfish end up where they shouldn't be, and what makes them so invasive? The answer to the first question is pretty straightforward: all signs point to the aquarium trade and fish hobbyists releasing the buggers, intentionally and/or accidentally. The first recorded lionfish on the East Coast was captured in Florida in 1985, and no more were reported until 1992. By 2010, they had reached the Gulf of Mexico including the coasts of Louisiana and Alabama. Since then the population has continued to skyrocket, leading to widespread habitat destruction and biodiversity loss, especially in the past couple of decades. Fascinatingly, genetic analysis found that all of the lionfish now in the Atlantic likely are the offspring of fewer than 10 initial females. That's quite a family tree!

As to what makes them so invasive, lionfish have a certain set of skills...

First: these fish can eat. A lot. They are non-selective feeders, meaning they eat pretty much anything they can fit inside them. Their stomachs can expand to 30 times its' normal size after a meal, so there's not much living near a reef that they can't squeeze in. They are known to eat fish that are 2/3rds their size. Their appetites border on gluttonous, having been recorded eating over 20 fish in a 30 minute period. Observations from North Carolina, and featured in this Slate piece, show well-developed layers of interstitial (visceral) fat, suggesting they are consuming far more than need and sustenance would call for. With an eating routine like that, it's no wonder these swimming digesters can wreak havoc on unsuspecting ecosystems. Even worse, native fish aren't used to being hunted by lionfish, so they don't recognize them as predators (a phenomenon known as prey naivete.)

Some examples of lionfish stomach contents, gathered in one day. Image Credit: Florida Fish and Wildlife
https://www.flickr.com/photos/myfwcmedia/

Second: if there's anything lionfish do as much as eat, it's make more lionfish. Most reef fish tend to spawn once a year (that's a very general statement.) Lionfish, however, reproduce year-round. I'm talking every three days. Females, who reach sexual maturity in less than a year, will continue to produce up to 30,000 eggs every three to four days for the rest of their lives. That's a potential 2 million plus offspring every year. With a lifespan believed to average around 10 years in the wild, and possibly longer (some have lived 35 years in captivity) they can quickly overwhelm native fish populations.

A baby lionfish in Duain, Philippines. Image Credit: Melissa Hobson, The Reef-World Foundation.
https://www.flickr.com/photos/reef-world/ 

There is one more factor that helps lionfish dominate these foreign ecosystems: they have no predators. In their home waters, it is believed that lionfish are eaten by large fish including groupers, snappers, sharks, and eels (though we don't actually know for sure!) However, being a fresh face on the Atlantic scene, no locals have developed a taste for the pillagers as of yet. It is certainly possible that some organisms may realize that lionfish are an untapped market and help to control their numbers, but until then, something else must be done to keep the swarms at bay.

The factors mentioned above are unfortunately not uncommon. They actually form a grouping of characteristics which can be found among many species which have ended up becoming invasive. The majority of invasive species, including the lionfish, will have at least a handful of these traits, and some organisms may even have the entire list. These traits include:

- Fast growth
- Rapid reproduction
- They move around/are carried a lot and/or quickly (high dispersal rates)
- Phenotype plasticity (they can often change to adapt to new environments)
- They can tolerate a wide range of environments and conditions
- Ability to survive on many food sources
- Association with humans (i.e. popular as pets, etc.)

Serving Up Solutions

Once the problem of lionfish in Atlantic waters began to be realized, it wasn't long before people got to work to try and mitigate the damage. Preventing the invasion in the first place is always ideal, but in this situation, it was far too late. Instead, people needed to find creative ways to start reducing lionfish numbers, and fast. When trying to control an out of control population, it's always prudent to enlist the help of some of the most successful hunters around: human beings.

Anglers were more than happy to lend a hand, or in this case a spear. Lionfish are notoriously difficult to catch with a hook and line, as they tend to avoid dead bait. That being said, there have been reports of some success using live shrimp. The go-to method is to stick them with a spear, or to catch them in a net. Most divers will carry a special canister or bag to store the caught lionfish safely and avoid getting stuck with a toxic spine. The storage space also makes each dive more effective, as going up and down repeatedly takes time and energy.

Several lionfish stored in a specially-designed bag during a dive. Image Credit: Florida Fish and Wildlife
https://www.flickr.com/photos/myfwcmedia/

Once the fishing community was onboard, the next step was to come up with something do with all the lionfish that were being caught. The simplest solution was to just eat them!

Hand-speared cooked lionfish served at a Dallas, TX restaurant. Image Credit: Dallas_Foodie
https://flickr.com/photos/funrun_9602/

Convincing people that eating an unknown and venomous fish is both safe and enjoyable is not easy. In order to drum up demand, groups in affected areas began a series of PR campaigns to educate the public about the problem of invasive lionfish and encourage them to chip in... with their stomachs. On May 16, 2015, the Florida Fish and Wildlife Conservation Commission (FWC for short) held the first annual Lionfish Awareness and Removal Day, which included a booth where guests could taste lionfish.

Guests gather to taste lionfish at the 1st Annual Lionfish Awareness and Removal Day, 5-16-15. Image Credit: Florida Fish and Wildlife
https://www.flickr.com/photos/myfwcmedia/

Things are going okay in terms of swaying public opinion, but the idea hasn't quite caught on yet at levels to be truly effective. New research from the U.S. Virgin Islands seems to suggest that the cost and availability of lionfish make it competitive with other readily available fish, and that public willingness to adopt them as a food source relates mostly to concerns over safety. As this article points out, part of the issue is confusion around what is venomous and what is poisonous. Venomous things are dangerous when you are injected by them (snakes, spiders, lionfish, etc.) while poisonous things are dangerous to touch or ingest (like some plants, toads, etc.) Lionfish are perfectly safe to eat when no longer living, and when the spines are removed.

Signs enlisting the public's help to "Be The Predator!" Image Credit: Florida Fish and Wildlife
https://www.flickr.com/photos/myfwcmedia/

Alas, even human hunting prowess is no match for the physical limits set by our corporeal husks. Even with a fully functioning SCUBA setup, the most skilled of divers can only safely go to depths of around 130 ft. The record for the deepest dive was set by an Egyptian named Ahmed Gabr in 2014 when he dove to a depth of 1,090 ft. (332.35 m) in the Red Sea. It took him 12 minutes to get down, but 15 hours to get back up, allowing proper time for safe decompression. What does this mean for the lionfish hunter?

Unfortunately, lionfish can inhabit a pretty wide range of depths. They can be found at depths of anywhere from 1 to 300 ft, meaning they can easily survive below the reach of humans. Lionfish are commonly caught in lobster traps set hundreds of feet deep, suggesting there is a sizable population constantly recharging the numbers of those caught in shallower waters. If this is the case, we will likely never effectively reduce populations by relying on human fishing alone. Luckily, we have some assistance from an unlikely ally!

Guardians of the Gulf-and-Keys

Imagine an underwater WALL-E mixed with RoboCop, searching the mean reefs of the west Atlantic for fishy offenders. This is essentially what a group called Robots in Service of the Environment is trying to accomplish. Their team of engineers, roboticists, biologists, and software programmers have designed a unique system with the sole purpose of seeking out and capturing lionfish in areas where they are considered invasive. The little bot is known as the Guardian LF1, and at the moment, is a Remotely Operated Vehicle (ROV) with plans to develop the system to become autonomous.

Analysis of the Guardian LF1 Mark 3. Image Credit: Robots in Service of the Environment
https://www.robotsise.org/the-guardian-solution/

Here's how the current system works, in a nutshell: a human on a ship controls the robot, which is attached to the ship through a cable (this sends communication signals as well as power.) The operator can then guide the Guardian around in any direction, including up and down, allowing them to seek out lionfish. When a target is spotted, the Guardian moves into position, and utilizes a small electrical shock to stun the creature. In a tribute to its' Roomba cousin (one of the co-founders of RSE was also a founder of the company iRobot, inventors of the Roomba,) it then sucks the stunned fish into a containment tube where they are safely stored until full. On each dive, the tank can hold about 30 lbs of fish, which amounts to about 10 lionfish on average. The Guardian is then brought to the surface, emptied, and redeployed instantly if needed.

A chart demonstrating the depth capabilities of the Guardian LF1. Image Credit: Robots in Service of the Environment
https://www.robotsise.org/the-guardian-solution/

This contraption provides some immediate advantages over human diving. For one, the robot can descend and ascend again immediately, with no need to wait for pressures to equalize. Human divers need to move slowly when returning to the surface to avoid decompression sickness (the bends.) The robot is capable of diving to depths out of reach of human divers, including the critical 200-500 ft range where the majority of lionfish breeding is believed to occur. The Guardian, being fed by a power cord, can keep on zooming until the operator gets tired of staring at the screen, and doesn't need to rest or take breaks. Another obvious benefit is that no humans are at risk of danger! By having the robot capture and handle the spiny fish, the chances of someone getting stuck by one are much lower. Scuba diving by itself is always at least a bit risky, even for those with plenty of experience under their belts, so any time an unnecessary dive is prevented, the chance of an accident occurring is reduced. On that note, learning to become a certified diver takes time, skill, and dedication. Essentially anyone can learn to operate the Guardians in a short period of time, and you don't need to know how to swim (you don't even need to get wet!)

There is certainly some room for improvement. Getting rid of the cord would be ideal but will obviously require a small yet powerful battery, as well as a wireless means of communicating with the controller and sending back visual data. This is easier said than done. Radio waves don't travel well under water, so most underwater vehicles use acoustics to transmit signals, which adds a couple seconds of delay for most systems. Seconds which may be critical when guiding a robot underwater. GPS tracking also utilizes radio frequencies, meaning underwater vehicles must also be tracked using acoustic positioning (known as LBL navigation.) If, or rather when these hurdles are overcome, the Guardian could theoretically be controlled from the comfort of ones living room. While one robot can certainly replace a few human divers in a single area, in order to effectively remove the need for human participation we would need a large fleet spread out over many different areas. This is certainly plausible, but the cost may be a deterrent (though likely cheap compared to what s total collapse of fisheries will cost.)

Learning To Swim

As work has continued on the Guardian's physical design, one breakthrough in particular has the potential to take this project to the next level: making the thing autonomous!

A still-image demonstrating the visual recognition system being tested on the Guardian LF1. Image Credit: Robots in Service of the Environment
https://www.robotsise.org/press-kit/

If you read my previous piece on the salmon cannon, you may remember it utilizes artificial intelligence software to sort between salmon and non-native or invasive species. The machine learning algorithms being developed for the Guardian LF1 are quite similar. The salmon cannon needs to ID fish quickly, as they pass through the tubing. The Guardian, on the other hand, can take its' time on making a decision. It can wait and allow the potential lionfish to move, or move itself to get a new angle, giving the algorithm more data to work with to spit out an ID. The salmon cannon also needed to differentiate between fish that may look very similar, and thus depended on calculations and ratios of the fish's body shape and size. Lionfish, luckily, are very unique in terms of shape and coloration, making differentiation by an AI program less of a challenge.

Once this software is up to snuff, the door will be open to self-driving autonomous underwater vehicles (AUVs) capable of searching out and capturing lionfish without human assistance. A control module (essentially a small ship) can be parked in the desired area allowing the Guardian to work it's magic. Luckily, research shows lionfish show high fidelity for their home turf, which is a fancy way of saying they don't move very far once they settle down. At a set interval, or potentially when the Guardian can find no more lionfish, the control ship can be relocated allowing the Guardian to search a new area. Ideally, the control ship itself would also be autonomous, allowing the whole system to be monitored and moved remotely. At some point, the stores of captured fish would need to be emptied, though this could easily be combined with refueling and other maintenance tasks. Just like that we have a relatively low-maintenance system which does the work of dozens of humans on its' own, and with minimal and remote human supervision.

Rise of the Robo-Conservationists

Necessity breeds innovation, and when it comes to environmental crises around the world, there certainly is no shortage of need for solutions. As robotic technologies improve, along with the AI and machine learning technologies that help them tick, more and more problems may be solved without direct human action. There are already some other examples of robots being used to help solve conservation and ecology challenges!

An inforgraphic outlining the COTSbot, developed by the Queensland University of Technology.
https://www.graphicnews.com/en/pages/33450/australia-starfish-killing-robot

Researchers at the Queensland University of Technology have developed an AUV known as COTSbot, which hunts and destroys COTS, or the crown-of-thorns starfish (Acanthaster planci) which have begun to devastate the Great Barrier Reef. Though the stars are native to the area, their populations have begun to grow unchecked, which scientists believe to be the result of overfishing of their natural predators as well as increased land-based nutrient run-off. When their numbers grow too large, they can overfeed on and kill coral reefs. COTSbot automatically navigates reefs, with the aid of cameras and sonar, and uses artificial intelligence to identify the starfish with an amazing  99.4% accuracy It then injects the target with a lethal dose of bile salts, which kill them without allowing them to separate and regenerate. The little bots move quickly (they can kill over 200 starfish in a single 4 to 8 hour mission) in an effort to outpace the rapid spawning of these echinoderms (a phylum containing members such as sea stars, urchins, and sand dollars.)

A red crown-of-thorns starfish feeding on coral. Image Credit: Matt Kieffer
https://www.flickr.com/photos/mattkieffer/

The potential of these systems are really exciting, and I'm eager to see how successful the implementation is. If it works well, we may begin to see a whole range of robots designed for specific conservation tasks. Even better, one could imagine a single robotic unit which could have different tools and modules attached depending on the needs of the work (like replacing the electric paddles and needles with nets, or shovels.) It could be that the Guardian and the COTSbot fail to bring the improvements we hope for, but it may also succeed beyond our greatest expectations. The only way to know is to try!

To dive deeper into the challenges posed by lionfish, crown-of-thorns starfish, and other invasive species, and learn more about how robots are being used to solve environmental issues, check out some of the following links and organizations:

- Lionfish.Co: An excellent site with tons of info and resources for learning about, or hunting, lionfish.

- Lionfish Central: A collaborative group working on finding lionfish solutions.

- National Invasive Species Information Center - Red Lionfish

- Robots in Service of the Environment: Designers of the Guardian LF1 robot.

- Sailors for the Sea - Red Lionfish