Everywhere you look, technology has become a part of modern day life. Science is renowned for its dependence on technology to generate vast quantities of data. But how effective is it really? Can technology actually replace the role of humans? My new study in Biological Conservation tries to solve the age-old dilemma of man versus machine by testing the effectiveness of camera technology as an alternative to human observers in a Peruvian small-scale shark and ray fishery.
Small-scale fisheries, despite their name, have an enormous impact on marine ecosystems. More than half of all fish consumed in developing countries originates from these fisheries with their name referring to the size of the fishing boats instead. Each boat can only harvest a small quantity of fish compared to industrial fisheries and relies heavily on manual labour, but the sheer number of boats in small-scale fisheries means the accumulative impact can be vast. Peru’s small-scale gillnet fishery alone consists of 3000 vessels, catching huge quantities of endangered species, including sharks, turtles, dolphins, seals and even penguins! Despite this, very little is known about their full impact as they are so difficult to study.
Cameras provide several advantages over human observers: they require little space on the already cramped vessels, they record data that can re-analysed by numerous independent analysts, and they have the potential for automation via artificial intelligence. Cameras also prevent the need of deploying someone at sea for long fishing trips under difficult and dangerous working conditions, and may ultimately save money from fewer salaries. However, there is the chance that cameras may miss some activity that an observer would not. Testing the effectiveness of these camera systems is therefore critical to understand the value of this data.
Our recent study installed cameras on five boats in Peru’s small-scale net fishery. This fishery targets sharks and rays, highly desirable and valuable fish worldwide, but also accidently catches numerous marine mammals, turtles and seabirds. Many of these species are endangered, meaning fishing activities could threaten their long term survival. Identifying areas where catch is high could help design new protected areas, whilst also improving our understanding of their current population.
Traditionally, the only way to monitor fisheries was the presence of on-board observers who recorded catch as it was hauled in. Use of observers in small-scale fisheries is severely restricted though, due to the sheer number of personnel needed to cover the fishery and the challenges of accommodating an extra person on-board these small boats. With the development of new, cheap technologies, remote electronic monitoring is becoming an increasingly popular tool to assess what impact these fisheries have on marine life. One potential tool is cameras.
Cameras generally performed well, detecting 9 of the 12 species of shark and ray captured on more than 90% of the time. Identification of incidental catch was more mixed. All cases of sea lion capture were recorded on camera, but only half of the sea turtle capture events were recorded on camera. It was possible though to identify these turtles to species level with 83% accuracy. Quantifying catch using cameras proved more difficult, with estimations highly dependent on the size, species and quantity of individuals captured.
Despite some deficiencies in the data generated from camera footage, the results are promising. Cameras have already shown their worth in identifying what species fishermen are catching, data that when combined with GPS can identify areas that these species inhabit. They also save considerable time and money when compared with the use of on-board observers, increasing the potential number of boats that can be monitored with the same budget. There is still much progress to be made, though, but with more and more developments to technology, it is likely the data will improve dramatically. This could help improve the conservation of numerous long-lived, charismatic and endangered species, which face the real prospect of extinction if action is not taken. It is surely only time before machine overcomes man, but for now, it appears that man is still needed in fisheries research.
Cameras provide several advantages over human observers: they require little space on the already cramped vessels, they record data that can re-analysed by numerous independent analysts, and they have the potential for automation via artificial intelligence. Cameras also prevent the need of deploying someone at sea for long fishing trips under difficult and dangerous working conditions, and ultimately save money from fewer salaries. However, there is the chance that cameras may miss some activity that an observer would not. Testing the effectiveness of these camera systems is therefore critical to understand the value of this data.
Our recent study installed cameras on five boats in Peru’s small-scale net fishery. This fishery targets sharks and rays, highly desirable and valuable fish worldwide, but also accidently catches numerous marine mammals, turtles and seabirds. Many of these species are endangered, meaning fishing activities could threaten their long term survival. Identifying areas where catch is high could help design new protected areas, whilst also improving our understanding of their current population.
The 12 different sharks and rays captured during the study. Images are from the cameras installed on the fishing boats.
In addition to the target species, Peru’s small-scale fisheries are prone to high levels of bycatch. During the study, dolphins, sea turtles, sea lions and penguins were all incidentally captured. Images are from the cameras installed on the fishing boats.
Bartholomew, D. C., Mangel, J. C., Alfaro-Shigueto, J., Pingo, S., Jimenez, A., Godley, B. J. (2018) Remote electronic monitoring as a potential alternative to on-board observers in small-scale fisheries. Biological Conservation, 219, 35-45. https://doi.org/10.1016/j.biocon.2018.01.003
Bartholomew Biology 