During the surveys ‘fish acoustics around Fernando de Noronha’ (FAROFA) performed in September 2017 (FAROFA1; https://doi.org/10.17600/18001399), April 2018 (FAROFA2; https://doi.org/10.17600/18001411) and April 2019 (FAROFA 3; https://doi.org/10.17600/18001381) a variety of optical methods were used to identify species and bottom habitat characteristics: vertical and towed GoPro video, stereo video system and a remotely operated vehicle (ROV). Vertical and towed GoPro video were captured with a GoPro TM Hero3+ operating in HD at 1,080 p equivalent to 2.1 megapixels per frame at 30 frames per second during FAROFA 1 and 2 and 60 frames per second during FAROFA 3.

• We achieved vertical video profiles with a camera fitted on a fishing pole equipped with an electric reel to film at different depths whilst adrift. Videos were captured with a GoPro TM Hero3+ housed in an aluminium enclosure Isotta CC-GoPro-B (FAROFA 1 and 2).
• We used a towed GoPro TM Hero3+ video on a downrigger along acoustic transects with the video looking down at the bottom (FAROFA 2 and 3).
• During FAROFA 3 survey, ROV video surveys were performed using a BlueROV2 system mounted with seven stainless ballast: two in the front and five in the back. The ROV was deployed with 150 m cable. Video-transects included soft bottom images. During four profiles, the ROV was equipped with a TDR sensor. Data were acquired through software Qgroundcontrol 3.5.0. To operate the ROV from the vessel, we used a joystick connected to a Microsoft Surface PC. Telemetry data were downloaded after recovering the ROV.
• During FAROFA 1 and 2 a Surface Video (GoPro TM Hero3+) was installed on the transductor metallic support and turned on, it recorded until the battery ran out.

Those video stations were located where noteworthy fish aggregations were observed on the echograms.

• Typical video profiles were 15 minutes long.
• Typical towed video were 50 minutes long.
• Videos were taken during day only

Each video were annotated using Solomon Coder free software to identify and enumerate observed species and sediment characteristics. Annotations contain fish species identified to specie level when possible and their maximum number observed. Benthic and pelagic fauna specimens were recorded from the images. A code was attributed to each specie consisting in the first three letters of the gender and first three letters of the specie separated by a point (Example: Canthidermis sufflamen code is can.suf). Information on sediment composition was recorded using different sediments categories. Presence of gelatinous and pelagic algae were also recorded and annotated in a qualitative way using “High”, “Medium” and “Low” values. Gelatinous abundance was underestimated for FAROFA 1 and 2 surveys.