Abstract
Twenty-four species of cetaceans (18 odontocetes, 6 mysticetes) regularly occur in the U.S. Exclusive Economic Zone of the Hawaiian Islands (Hawaiian EEZ). Abundance estimates are needed to evaluate the impacts of human activities in population assessments of these species. The Hawaiian Islands Cetacean and Ecosystem Assessment Survey (HICEAS) is a recurring ship-based, line-transect survey designed to estimate cetacean abundance in the entirety of the Hawaiian EEZ. Given the vast study area, two ships operating a total of approximately 180 days within the summer-fall period are required to complete each HICEAS. To date, HICEAS has been conducted in 2002, 2010, and 2017. Low encounter rates in the study area require that sightings of the same and similar species be pooled with sightings from previous line-transect surveys when estimating detection functions. Thus, estimating cetacean abundance during HICEAS 2017 offered an opportunity to update abundance estimates from HICEAS 2002 and
2010 using the most current detection functions and new estimates of trackline detection probabilities that consider the effect of survey sighting conditions. Group size and Beaufort sea state were the most important factors affecting the detectability of cetacean groups. Abundance was estimated for 21, 19, and 18 species in 2002, 2010, and 2017, respectively, with 16 species (14 odontocetes, 2 mysticetes) accounted for in all HICEAS years. Across all species and years, abundance point estimates range from 137 blue whales (Balaenoptera musculus) in 2010 to 76,375 rough-toothed dolphins (Steno bredanensis) in 2017. The low encounter rates led to high CVs (range, 0.27 to 1.71) for most estimates and low power to detect trends in abundance during the study period. Additionally, random variation in the sampling process and sighting attributes, along with interannual variation in oceanographic conditions within the Hawaiian EEZ, had pronounced effects on the abundance estimates, further complicating comparisons among years. Habitat-based modeling, satellite tagging, photo-identification, acoustic analyses, and simulation approaches can provide additional temporal and spatial inference that may be needed to assess and manage high priority species.