Rules of attraction: enticing pelagic fish to underwater cameras

One aspect of my PhD is to understand a little bit more about the ecology of pelagic fishes, specifically what habitats they prefer in coastal waters. But before I could do this I first had to determine the best way to sample these fishes and this is what todays post is on.


Incredible painting of a Blue Marlin by the talented James Prosek.

Pelagic fish are those charismatic species such as Yellowtail Kingfish, Marlin, Bonito, Tuna, Australian Salmon, etc. that occupy the mid to upper levels of the water column*. Compared to demersal fish (fish that live close to the seafloor), very little is known about pelagic fishes in coastal environments. This is primarily due to difficulties in surveying pelagic fish numbers and their patterns of distribution. Pelagic fish are really fast swimmers so they area capable of avoiding conventional survey gear like nets. In coastal settings they live in challenging habitats, which are often deep, wave exposed and adjacent to rocky headlands. And to top it off they are very sporadic in their occurrence in time and space. For example, I’ve spent many afternoons fishing for pelagic fishes catching zilch only to return the following day and catch a fish every cast!

The challenges associated with sampling pelagic fishes have resulted in limited ecological research on pelagic fish in coastal ecosystems. This is somewhat concerning as they play an important role in near-shore food webs and are heavily exploited by commercial and recreational fishers. In fact, for some species (e.g. Bonito) there are barely any studies on things like their basic biology and stock status.


Australian Bonito caught spinning from the rocks. PC Matt Rees.


Ian Osterloh with a beaut Yellowtail Kingfish caught on a surface popper. PC Ian Osterloh.

A potential sampling technique to overcome the issues of surveying pelagic fishes is using Baited Remote Underwater Video Systems (BRUVS). BRUVS are a video camera inside an underwater housing attached to a steel frame or bar. The camera looks out over bait inside canister (usually crushed pilchards) which attracts fish into the camera’s field of view. BRUVS are generally deployed between 30 – 60 minutes. The footage is then reviewed back in the office to determine what species of fish and how individuals of each species are in a particular area.

Researchers have found BRUVS positioned in the mid-water with 500-1000g of bait to be a useful tool to sample pelagic fishes (see this video and these papers here, here and here). However, these studies still have issues due to the patchy occurrence of pelagic fish turning up to underwater cameras. For me this raised the question whether there are other attractants, or a combination of attractants that can gain better estimates of coastal pelagic fish populations.

Considering the biology of pelagic fish, many of which display schooling behaviour and are fish-eating predators, I thought that attractants associated with sound and sight stimuli may offer potential alternatives. So in the Jervis Bay Marine Park on the south coast of NSW I ran an experiment to compare the relative importance of sound, sight and scent in attracting pelagic fishes to underwater cameras.



The study location; Point Perpendicular in Jervis Bay Marine Park. This stretch of coast is one of NSW’s best game fishing areas. You can see land based fishers setup at ‘the tubes’ on the left in the pic.

To do this I set up 5 mid-water video cameras each with a different ‘attractant’ treatment. The first camera had a scent attractant, which was 500 grams of crushed pilchards. The second camera a sound attractant, which was a speaker located above the camera playing baitfish noise (the sound of Yellowtail Scad and Blue Mackerel swimming and feeding). The third camera had a sight attractant (reflective material attached to the camera). The fourth camera had scent, sight and sound attractants attached to examine interactive effects. Finally the last camera was a control which had no attractant.

Over a series of days, weeks, months I deployed these cameras 18 times along Point Perpendicular (one of Australia’s best game fishing areas) in 20 metres of water 50 metres from the shore. I then looked through all the footage counting the number of pelagic fish I observed between the 5 treatments. Checkout the highlights video below to see some of the species we recorded and click 1080 HD :).


Deploying a mid-water camera.


The sound equipment. The underwater speaker is positioned just below the surface of the water. Amplifier and 12 volt battery is housed on the surface in an esky.


Despite a lot of variability, we found that cameras with all attractants (scent*sight*sound) recorded a substantially greater number of pelagic fishes compared to the cameras with a single attractant or no attractant. The average number of fishes recorded on the camera with all attractants was 100 individuals, whereas the cameras with a single attractant or no attractant recorded on average 15 individuals or less. This pattern was primarily driven by two species; Yellowtail Scad and Australian Bonito. Interesting bait, the standard attractant used in BRUV studies performed poorly in enticing pelagic fishes to our underwater cameras. We also found that the arrival time of pelagic fish to the camera with all attractants was significantly shorter compared to the cameras with one or no attractant (17 vs. 31 min, see figure 1 below). So if you want to attract pelagic fishes to underwater videos our findings indicate that you need to use multiple attractants!!!


Arrival time of pelagic fishes to the camera with all attractants attached (Sight*Sound*Scent) versus the cameras with a single or no attractant attached.

But why would the combination of sight, sound and scent attractants interact to entice and record a greater number of pelagic fish? Well we propose that the effect of multiple attractants could be the result of higher detectability of the camera by fish over a larger area. For instance, sound travels 5x faster in water compared to air and propagates in all directions equally. Therefore, sound may be attracting fish over a broad area. Bait is likely to be effective attractant up to 200 metres based on previous research, while visual stimuli up to 50 metres. Therefore, we speculate that fish first hear the sound recordings, which guides them until they detect the bait plume. The visual stimuli then attract the fish closer to the camera.

In conclusion, our findings show that it’s important to consider attract type when surveying coastal pelagic fish populations. We suggest future studies use multiple attractants when surveying pelagic fish with underwater cameras.

by Matt Rees.

You can find the full paper here:

Rees, M. J., Knott, N. A., Fenech, G. V., & Davis, A. R. (2015). Rules of attraction: enticing pelagic fish to mid-water remote underwater video systems (RUVS). Marine Ecology Progress Series, 529, 213.

A huge thanks to NSW Department of Primary Industries, The Nature Conservancy Australia and the Ecological Society of Australia for supporting this research.

NC  dpi

ESA logo                Basic CMYK


*Note: some of these species may be considered transient between demersal and pelagic environments as they are also associated with the seafloor. For the purposes of this study I treat them as ‘pelagics’ as I was sampling in the mid-water.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s