Inferring trackline detection probabilities,g(0), for cetaceans from apparent densities in different survey conditions

Visual line‐transect surveys are commonly used to estimate cetacean abundance. A key parameter in such studies is g(0), the probability of detecting an animal that is directly on the transect line. This is typically considered to be constant for a species across survey conditions. A method is developed to estimate the relative values of g(0) in different survey conditions (Beaufort state) by comparing Beaufort‐specific density estimates. The approach is based on fitting generalized additive models, with the presence of a sighting on a survey segment as the dependent variable, Beaufort state as the key explanatory variable, and year, latitude, and longitude as nuisance variables to control for real differences in density over time and space. Values of relative g(0) are estimated for 20 cetacean taxa using 175,000 km of line‐transect survey data from the eastern and central Pacific Ocean from 1986 to 2010. Results show that g(0) decreases as Beaufort state increases, even for visually conspicuous species. This effect is greatest for the least conspicuous species (rough‐toothed dolphins, beaked whales, minke whales, and dwarf and pygmy sperm whales). Ignoring these large effects results in a nontrivial bias in cetacean abundance estimates.