What Fitbit tech and algorithms can reveal about lemon sharks March 13, 2018 Cryptic lives: Accelerometer technology is helping researcher Dr Lauran Brewster (pictured) to shed light on the behaviour of juvenile lemon sharks. Murdoch University researchers have used Fitbit-like accelerometer technology and an algorithm to understand more about the behaviours of wild juvenile lemon sharks. In a study led by postdoctoral researcher Dr Lauran Brewster, the data revealed this near threatened species eats more food during the wet season than the dry season, during low tides when there are less predators and during the early evening. Dr Brewster said it was important to understand the behaviour of sharks as they play an important role in balancing their ecosystem, often occupying top predator positions. “The data is helping us to unveil the cryptic lives of these animals, showing us what they do on a daily basis,” she said. “Accelerometers – like those used for monitoring your movement in a Fitbit – allow us to look at the body movement of wild animals in their natural environment. Normally it is very difficult to study them there. “The accelerometers record huge amounts of data so we needed to teach a computer an algorithm to classify patterns within the data into behaviours we witnessed juvenile lemon sharks performing, including catching prey.” Lemon sharks are often found in the shallow subtropical waters of the Americas and West Africa. They are classified as Near Threatened by the IUCN (International Union for Conservation of Nature). Dr Brewster and her research team discerned five behaviours performed by juvenile lemon sharks fitted with accelerometers in the Bahamas. At first, sharks in captivity were observed while wearing accelerometers. They exhibited a range of behaviours including burst swimming, ‘head shaking’, chafing (where a shark is swimming and will roll to ‘scratch its back’ on the sand), resting and steady swimming. These observations were then used to train computer algorithms to recognise unique patterns in the accelerometer data associated with these behaviours, in order to match them to the data retrieved from the accelerometers that were fitted to sharks in the wild. The ‘head shaking’ behaviour was observed when sharks caught prey, so this method helped researchers to understand when this was happening in the wild. Dr Brewster said being able to identify important behaviours for an animal, their frequency and the habitat in which they occur, gives scientists an indication of important areas for protection and management. “If we can gain an understanding of an animal's daily activities in pristine environmental conditions then these can be compared to those of their counterparts in degraded environments to see what the impacts are and possible ways to mitigate them.” The lemon sharks and accelerometers project is being led by Dr Adrian Gleiss from Murdoch’s Centre for Fish and Fisheries Research and Dr Nicholas Whitney from New England Aquarium in the United States. The research is a collaboration between scientists from a number of international institutions including the Bimini Biological Field Station Foundation in the Bahamas, the University of Hull in the UK, USA institutions Stanford University, the Rosenstiel School for Marine Science and Technology and the University of Massachusetts. Their paper, entitled Development and Application of a Machine Learning Algorithm for Classification of Elasmobranch Behaviour from Accelerometry Data has been published in Marine Biology. It can be viewed here. Print This Post Media contact: Jo Manning Tel: (08) 9360 2474 | Mobile: 0408 201 309 | Email: email@example.com Categories: General, Research, Animal and plant studies, environment and bioinformatics, School of Veterinary and Life Sciences Research Tags: accelerometers, accelerometers and sharks, adrian gleiss, animal behaviour, centre for fish and fisheries research, fish research murdoch, fitbits and sharks, iucn, lauran brewster, lemon sharks, shark research Leave a comment Name (required) Mail (will not be published) (required) Website You can use these tags : <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong> We read every comment and will make every effort to approve each new comment within one working day. To ensure speedy posting, please keep your comments relevant to the topic of discussion, free of inappropriate language and in-line with the editorial integrity of this newsroom. If not, your comments may not be published. Thanks for commenting!