MEPS 197:259-272 (2000)  -  doi:10.3354/meps197259

Baited technique improves censuses of cryptic fish in complex habitats

Bryce D. Stewart*, Joanne S. Beukers

James Cook University, Townsville, Queensland 4810, Australia
*Present address: Port Erin Marine Laboratory, University of Liverpool, Port Erin, Isle of Man IM9 6JA, United Kingdom. E-mail:

ABSTRACT: Many fish species are cryptic by nature or show a negative response to divers. This may make traditional censusing techniques difficult to perform and the results obtained using these methods may be inaccurate. This problem will be exacerbated in complex habitats, such as those found on coral reefs, because the cryptic species present will be even more difficult to locate and identify. Predation studies on coral reefs have been especially hampered by this problem because many predatory species are cryptic and so it has been difficult to obtain reliable abundance estimates. This study tested a visual census technique that used bait to bring cryptic predatory fish into view. Results from this census were then compared to those from a traditional survey using belt transects, and to a patch reef tagging study where all individuals of 3 key species were located. The baited technique produced significantly higher density estimates for 3 of the 4 most abundant cryptic species. The patch reef experiment demonstrated that the baited technique accounted for 85 to 96% of fish present. Censuses without bait observed only 40 to 61% of fish present. For mobile species, on the other hand, the baited census appeared to overestimate abundance, due to movement of fish into the census area. We therefore recommend combining baited censuses of sedentary cryptic species with traditional censuses of mobile species to gain an accurate picture of piscivorous reef fish communities. Using this approach at Lizard Island on the Great Barrier Reef, the proportion of cryptic piscivores in the community was found to be almost double that which would have otherwise been observed. We also found considerable spatial variation in the abundance and distribution of piscivorous fish. These patterns would have been quite different had the survey been based on belt transects alone. Previous studies using belt transects may therefore have underestimated the importance of cryptic piscivorous species in communities of coral reef fish.

KEY WORDS: Visual censuses · Baited censuses · Cryptic species · Piscivorous fish · Predation · Coral reefs · Community structure · Spatial variation

Full text in pdf format