Shark Smarts: Can associative learning change shark behavior?

Reference: Mitchell, J. D., et al. “A novel experimental approach to investigate the potential for behavioural change in sharks in the context of depredation.” Journal of Experimental Marine Biology and Ecology 530 (2020): 151440.

Can sharks learn?

North West Cape, Australia. This cape is a peninsula, so is surrounded on three sides by water, and contains a coral reef in addition to a no-take marine reserve. Image by Murray Foubister via Wikimedia Commons.

Most of us are familiar with nature documentaries involving sharks. They often include researchers on a boat throwing chum or bait into the water in an attempt to lure sharks for tagging or observational purposes. This is clearly an effective way to attract sharks and study them in their natural habitat, and it is a far easier method than researchers scouring areas where sharks dwell in the hopes of spotting one. However, through all of these bait-deployment studies, researchers have never really examined if this act of luring sharks repeatedly alters their behavior. Are sharks more likely to swim up to research vessels in areas where bait-deployment studies are repeatedly used? Does the presence of boats make sharks more likely to feed from these bait-deployments? Can sharks learn, over time, that a boat putting out bait means food for them, and act accordingly? These are questions a research team located in western Australia tried to answer.

Developing a bait-deployment system

Researchers chose four study sites in North West Cape, Australia. Two of the sites were in a no-take marine reserve, which are areas where no fishing is allowed to occur. These areas exist across the globe, and are meant to aid in fish conservation. The other two study areas were outside of the no-take marine reserve, where many fishing boats were present.

In order to entice sharks to approach the research vessel, a weighted metal rod that had a bait bag full of crushed fish, as well as two salmon heads attached to it, was lowered into the water at the study sites. A float was attached to the top of the rod to prevent it from sinking to the bottom, and two downward facing cameras were affixed to the rod to ensure sharks were documented as they approached the bait bag and salmon. Researchers released the bait rod three times a day over six days, and quantified the time of arrival for a shark, the time of first feeding, as well as the species of shark present.

An illustration of the bait-deployment system used by researchers in the study. In addition to what is shown, the boat was anchored to the seafloor. Recreated from Mitchell et al. 2019 by Francesca Giammona. Image created in


How did the sharks behave?

A sickefin lemon shark, the most common shark seen during the study. These sharks were so common that 33 were seen during the six day study period out of 63 sharks total. Image by Patrick Quinn-Graham via Wikimedia Commons.

After the six days, there were very clear trends that the time to arrival and time to first feeding for sharks had decreased rapidly, occurring about three times faster at Day 6 compared to Day 1 for both measurements. Four different shark species were observed feeding at the study sites in the fishing areas; the most common species seen feeding was the sicklefin lemon shark. In contrast, only two species approached the bait rod in the marine reserve areas, and these species did not feed. Tiger sharks were the most common in these no-take marine reserve areas. There was a lot of variation between species in regards to the amount of time each took to approach the bait rod and then feed, which indicates that all shark species do not learn or interact with bait in the same way. Overall, blacktip sharks were the fastest to approach the research vessel and feed, while tiger sharks were the slowest to approach.

It was clear that repeating the bait rod deployment over consecutive days had an effect on the sharks. It was even observed that in the later days of the experiment, sharks took less than one minute from the time the bait rod entered the water to feed. The researchers believe this is an example of associative learning, where the sharks associate the research vessel with food, and swim near it until that food becomes available as the bait rod is deployed. This learning likely relies heavily on smell, as the smell of the bait initially drew sharks to the site, but this learning likely also relies on sound. As the research vessel moved through the water, it made a clear noise via the propeller that the sharks could follow, once they learned that the vessel indicated food. This would allow them to feed immediately once the bait was available, as they were already near the research vessel. It is possible that the sharks in the area had already associated the sounds of typical fishing vessels with food, as these boats also throw out bait often. This could explain why so many more feedings occurred outside the marine reserve, where fishing vessels are free to roam, rather than inside the no-take marine reserve, where no fishing usually occurs and the sound of a boat would be unfamiliar to the sharks that usually swim there.

Why does this matter?

This study shows a clear instance of associative learning in sharks, and can help to inform experimental set up for shark researchers who want to study these creatures using a bait-deployment system. It brings to light that sharks can not only learn to associate sensory cues (i.e. sound) outside of what they may normally experience with food, but that each shark species may learn differently in different environments (i.e. in an area with fishing versus an area with no fishing). This work can also educate fishermen who work in these waters, where sharks are common, to the fact that the act of fishing may draw more attention from these predators over time. For the rest of us, this new knowledge about sharks can hopefully shape our mental image of them from one of a predator ruled by only instinct to one that is capable of behavioral change, even after being on the planet for 450 million years.

Leave a Reply

Your email address will not be published.