A statistically robust sampling design must be chosen, allowing for adequate spatial/temporal coverage and replication whilst meeting the overall survey objectives, given available equipment and vessel time. Santana-Garcon et al. (2014b) recommend a minimum of 8 replicates per experimental treatment in warm-temperate and tropical coastal environments, although this may be dependent on the geographic distribution and abundance of species. The final design should be communicated to all personnel before the survey to maximise clarity and efficiency during field operations. As a rule, pelagic BRUVs should be deployed a minimum of 200-500 m apart to reduce the likelihood of bait plume overlap and inter-camera animal movements (Santana-Garcon et al. 2014b, Bouchet & Meeuwig 2015), but further field testing is required to determine if this separation is sufficient to consistently guarantee independence between replicates when sampling large, mobile vertebrate species. See Chapter 2 for additional details.

The timing and duration (“soak time”) of BRUV deployments should be determined. Deployments conducted 30-60 min after sunrise and before sunset should abate the effects of differential crepuscular behaviour in fishes (Axenrot et al. 2004, Potts 2009). If BRUVs are only one part of a larger research programme, it is important to think carefully about the timing of BRUV operations, as bait use may bias subsequent observations at that same site (e.g. if diver surveys were to follow). Optimal soaking time is likely to vary across habitats and represent a practical compromise between increasing sample size and making the best use of available vessel time in light of the target level of replication. Previous studies have reported soaking times of 45 min (Rees et al. 2015), 120-135 min (Letessier et al. 2013b, Santana-Garcon et al. 2014c), 165 min (Bouchet & Meeuwig 2015), or 180 min (Santana-Garcon et al. 2014b). Santana-Garcon et al. (2014b) suggested a soak time of 120 min. In cool-temperate waters, Bouchet & Meeuwig (2015)’s species accumulation curves failed to plateau after 3 hours. Although some attempts have been made to develop a range of plausible bait plume dispersal models (e.g. Olsen & Laevastu 1983; Sainte-Marie & Hargrave 1987), further on bait diffusion in the mid-water is needed to confirm the minimum distance that should be allowed between deployments, estimate the effective sampling area in a range of conditions, and better understand the dynamics bait flushing across different levels of fish activity. Lastly, careful thought must be given to the choice of suspension depth, as different assemblages may vary along depth gradients away from the surface (Santana-Garcon et al. 2014b).

Consideration must be given to the location of BRUVs during deployment. Instruments should not be deployed where there is a risk of entanglement (e.g. near fishing gear) or where they are likely to constitute or become a navigational hazard (e.g. inside shipping lanes, where trawlers are operating). At a minimum, deployment and retrieval locations should be recorded, with vessel location monitored at regular time intervals as a back-up. GPS loggers can be mounted on flag poles or buoys when deploying free-drifting BRUVs and are advised for capturing the exact spatial trajectories of the units (Bouchet & Meeuwig 2015). VHF radio beacons are also recommended to avoid gear loss in adverse weather conditions. Geofencing technology could be used (as it has been with fish aggregation devices) should the user need to be alerted when BRUVs exit a predefined area.

Appropriate approvals must be obtained. All research activities within Australian Marine Parks are to be undertaken under permit, and most institutions will also require Animal Ethics approval, even if the proposed methods are non-invasive. All institutional health and safety requirements must also be satisfied (e.g. travel risk assessment, volunteer insurance proposal). See Appendix B for a list of potential permits required at the Commonwealth level).

Appropriate camera settings must be selected (e.g. frame rate, video resolution, field of view mode, action cams vs camcorders, see Table 6.3) in light of their performance relative to the study goals and market availability. Correct date/time settings are particularly crucial for file management during subsequent analyses. When using GoPro cameras, note that standard and dive housings are rated to 40 m and 60 m respectively. Special backdoors must be also fitted if battery packs are considered. All equipment must be carefully checked prior to deployment, including that cameras have been serviced, cleaned, and calibrated (if using stereo-BRUVs). Spares (batteries, memory cards, cameras, Table 6.2) are essential as a contingency plan against equipment failure/damage/loss or adaptive changes in the sampling plan (e.g. additional deployments).

Table 6.3: Example camera settings for a pelagic BRUVs. Values reflect the use of GoPro Hero3 cameras. Options may differ in other camera models.

Settings Value
Resolution 1080
Frame Rate 25 fps
Field of View Medium
Upside Down Up
Spot Meter Off
Looping Video Off
Set up
Default Mode at Power Up Video (default)
One Button Off
Onscreen Display ON
Camera Status Lights 2
Sound Indicator Off
Manual Power Off Manual

Bait must be ordered ahead of time in sufficient quantities. Sourcing bait locally from factory discards (e.g. fish heads, tails and guts) is an attractive option for reducing costs and the ecological footprint of sampling. For some applications, bait balls comprising minced fish, oil, and/or meal, may also be appropriate, though care should be taken to standardise bait mixtures across deployments. Between 800g-3kg of bait is generally adequate for deployments of up 3 hours (Letessier et al. 2013b, Santana-Garcon et al. 2014b), though having extra supplies (e.g. 20%) may be useful if extra/longer deployments can/must be undertaken. Ultimately, the choice of bait quantity should be informed by consideration of the desired soaking time, expected flushing rate, and likely level of fish activity. Sufficient freezer space must be made available on-board accordingly. Debate is still ongoing over the most efficient way to prepare bait, although crushed/slurried mixtures seem more likely to disperse well into the water column. Presentation is also important, with wire mesh baskets (Santana-Garcon et al. 2014b) and perforated PVC tubes (Bouchet & Meeuwig 2015) being two popular options, despite the lack of comparative studies of their relative efficiencies. Critically, recent research demonstrates that bait alone may be a biased/poor attractant for pelagic fishes, and that consideration should be given to combinations of multiple attractants associated with sight, sound, and scent to help generate more effective abundance estimates for some species (Rees et al. 2015).

Rig set up should reflect the chosen BRUV design, and may need to be adapted in response to vessel constraints (e.g. available deck space). It is critical to check that the correct amount of weight, length of ropes, number of buoys etc. are available before the survey begins (Figure 6.1). Spare units and parts are essential in all circumstances.

Sampling gear specifications should always be fully documented to achieve maximum transparency and comparability. Over a third of studies fail to report on basic methodological choices (Whitmarsh et al. 2017), including rigging plans, camera orientation, spacing, convergence angle, field of view, inter-BRUV distances, soak time, bait choice and quantity, bait preparation technique, bait dispenser type, suspension depth, deployment configuration (Figure 6.1), number of replicates, among others.

Data storage needs must be anticipated. 2TB portable hard drives will typically provide enough storage space for 100 hours of high-resolution video footage, though this may vary by camera model/make. Equally important is making sure that enough power boards, adapters, USB hubs, data cables, etc. are purchased, and can be configured safely for use at sea, so that data offload and backup following each deployment can occur. Planning for double copies of each hard drive and for offline storage on institutional servers is highly recommended to avoid data loss in the event of hardware failure.

Pre-survey checklist

Task Description/comments
Sampling design chosen and coordinates of sampling sites calculated and checked for safety hazards
Pelagic BRUV design and configuration determined
Deployment protocol determined, including methods for locating/tracking gear
Appropriate permits obtained and printed copies made (on waterproof paper if necessary)
Bait (and/or other attractants) ordered in adequate quantities
Camera settings determined, and cameras calibrated as appropriate
Data storage needs identified and hardware purchased accordingly
Metadata sheet prepared
Gear shipment arranged