Effect of underwater visual survey methodology on bias and precision of fish counts: a simulation approach

• DOI: 10.7717/peerj.5378
• Publication Date: 2018-07-30T08:10:56Z
• Authors: Miguel Pessanha Pais; Henrique N. Cabral
• Keywords: Sampling bias

Bias in underwater visual census has always been elusive. In fact, the choice of sampling method and behavioural traits fish are two most important factors affecting bias, but they still treated separately, which leads to arbitrarily chosen methods. FishCensus, a two-dimensional agent-based model with realistic movement, was used simulate problematic SCUBA diving methods understand how methodology affects precision bias counts. Using fixed true density 0.3 fish/m2 visibility 6 m, 10 counts were simulated for several combinations parameters transects (length, width, speed) point (radius, rotation speed, time), generating trait-specific heatmaps precision. general, had higher less precise than transects. Fish attracted divers led highest while cryptic accurate For counts, increasing survey time increased variability, radius reduced case that avoid divers. Rotation speed did not have significant effect it Wider longer faster swim beneficial when mobile species, narrower, shorter transect slow is fish.