Bay watch: Using unmanned aerial vehicles (UAV's) to survey the box jellyfish Chironex fleckeri

• Published in: PloS one 2020-10, Vol.15 (10), p.e0241410
• Main Authors: Rowley, Olivia C, Courtney, Robert L, Browning, Sally A, Seymour, Jamie E
• Format: Article
• Language: English
• Subjects: Animals; Autonomous underwater vehicles; Biology and Life Sciences; Body Size; Box jellyfishes; Chironex fleckeri; Cloud cover; Clouds; Computer and Information Sciences; Cubozoa - anatomy & histology; Drones; Earth Sciences; Ecological effects; Ecology and Environmental Sciences; Environmental conditions; Environmental Monitoring - instrumentation; Hydrographic surveying; Marine technology; Medicine; Methods; Observations; Polls & surveys; Population Dynamics; Remote sensing; Species; Success; Surveys and Questionnaires; Target detection; Technology; Unmanned aerial vehicles; Wind speed; Zoological research
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0241410

Biological investigations on free ranging marine species are regarded as challenging throughout the scientific community. This is particularly true for 'logistically difficult species' where their cryptic natures, low abundance, patchy distributions and difficult and/or dangerous sampling environments, make traditional surveys near impossible. What results is a lack of ecological knowledge on such marine species. However, advances in UAV technology holds potential for overcoming these logistical difficulties and filling this knowledge gap. Our research focused on one such logistically difficult species, the Australian box Jellyfish (Chironex fleckeri), and we investigated the capacity of consumer grade UAV technology to detect this, highly venomous, target species in the inshore waters of Northern Queensland Australia. At two sites in the Weipa area, we utilized video analysis, visual count comparisons with a netted animal tally, and evaluated the role of associated environmental conditions, such as wind speed, water visibility and cloud cover on jellyfish detection rates. In total fifteen, 70 meter transects were completed between two sites, with 107 individuals captured. Drone success varied between the two sites with a significant difference between field and post-field (laboratory) counts. Animal size and cloud cover also had significant effects on detection rates with an increase in cloud cover and animal size enhancing detection probability. This study provides evidence to suggest drone surveys overcome obstacles that traditional surveys can't, with respect to species deemed logistically difficult and open scope for further ecological investigations on such species.