Detectability of Pacific crown-of-thorns starfish (Acanthaster cf. solaris) using scooter-assisted large area diver-based (SALAD) surveys in the Northern Great Barrier Reef, Australia

Abstract Limited detectability of crown-of-thorns starfish (CoTS) undermines the capacity to survey and manage population irruptions, though few studies have considered the factors that influence detectability. This study explores spatiotemporal variation in detectability of CoTS during emerging population irruptions in four regions of the Northern Great Barrier Reef (GBR), by comparing recorded densities based on the number of CoTS sighted to inferred densities that account for distinct sets of feeding scars, which are presumed to indicate the localised occurrence of starfish that cannot be detected. Mean detectability was estimated to be 50.51% (± 1.59SE) across all scooter-assisted large area diver-based (SALAD) surveys where CoTS or feeding scars were recorded ( n = 405 surveys), ranging from 29.59% (± 4.22SE) up to 59.46% (± 2.69SE) among regions. Regional differences in detectability correspond with differences in the size of CoTS, whereby larger CoTS tended to be more exposed, and marked differences in the proportion of CoTS recorded feeding. The average exposure increased with size of CoTS, with an asymptote at approximately 60% for CoTS > 500 mm diameter. Proportional exposure of individual CoTS was also strongly affected by feeding, whereby CoTS that were feeding during the day were generally completely exposed. This study reaffirms that limited detectability of CoTS needs to be taken into account when estimating absolute densities using different survey methods. These data also highlight significant limitations of survey methods (e.g. manta tow and towed camera) that rely on visual detection from above to estimate CoTS densities.