LAUSR

Abstract Gear performance is often assumed to be constant over various conditions encountered during sampling; however, this assumption is rarely verified and has the potential to introduce bias. We used fishery-independent vertical line surveys to evaluate whether gear efficiency and selectivity is similar while assessing reef fish populations at oil and gas platforms, artificial reefs, and natural banks in the western Gulf of Mexico. We conducted 192 vertical line sets with cameras placed on a subset of these deployments to validate any differences in efficiency among habitat types. Red snapper (Lutjanus campechanus) accounted for 93% of the catch. No difference in red snapper CPUE among habitats was detected. When evaluating fish size, 8/0 and 11/0 hooks sampled significantly larger red snapper at natural banks than at artificial habitats. While CPUE was similar among all hooks at artificial habitats, CPUE at natural banks was lower for shallower hooks and increased towards the bottom hooks along the backbone. At all habitats, red snapper TL decreased from shallow to deep hook positions. Simultaneous camera deployments revealed other processes affecting efficiency such as bait removal and depredation. Vermilion snapper (Rhomboplites aurorubens) were effective at removing bait while avoiding capture. Perhaps related to this observation, Red snapper CPUE was negatively correlated with the vermilion snapper video index of abundance. Video confirmed gear saturation was prevalent (70% of deployments), occurring more frequently on artificial habitats. Furthermore, the time fished was effectively “shorter” at artificial habitats as the number of available baited hooks declined rapidly. These results point towards higher relative abundance at artificial habitats; however, the prevalence of saturation indicates vertical line CPUE may not always be proportional to true abundance, hindering our ability to detect differences at the scale examined in this study. Vertical line surveys should evaluate the prevalence of saturation as inferences regarding relative abundance may be compromised when this information is unknown.