612 



Flshen/ Bulletin 101(3) 



ever, there were too few sightings to make ^^(0) estimates 

 because the overall group encounter rate was lower than 

 anticipated and the CIO was only used for 35% of the total 

 survey effort. In the future, a CIO should be used whenever 

 the primary team is on-effort and the CIO should search an 

 area larger than 30° left and right of the bow. Although the 

 data in proximity of the transect line are most critical for 

 estimating ^(0), it is also necessary to have enough data to 

 estimate fiO) for groups missed by the primary team. 



More work is needed to develop methods for estimating 

 g(0) in relation to perception bias in the southern U.S. At- 

 lantic. Completely independent observers cannot be used 

 because the ship has to be diverted from the transect line 

 to identify species and make group-size estimates. Because 

 many groups can easily be lost once sighted, the ship must be 

 diverted well before the group passes abeam. Barlow ( 1995) 

 used a CIO that searched the same area as the primary team 

 with unaided eye or 7x binoculars. The 25x binoculars were 

 used in our study to increase the number of CIO sightings 

 and avoid attraction bias in/tO). Previous experience in the 

 Gulf of Mexico has indicated that many unaided-eye sight- 

 ings would be of small groups of species that are attracted 

 to the ship to ride the bow waves. Conversely, small groups 

 are the most difficult for an independent observer to track 

 with 25x binoculars because the ship is not diverted and the 

 bearing to the group is constantly changing. 



Similar to Barlow's (1995) findings on perception bias, 

 the majority of groups missed by the primary team were 

 apparently small groups, although the group-sizes were 

 not estimated at close range. Barlow (1995) estimated g'(O) 

 ranging from 0.73 and 0.79 for small groups of delphinids 

 (<21l and cryptic species (which usually occur in small 

 groups), andg(O) = 1 for groups of >20 delphinids. In ad- 

 dition to group-size, the magnitude of perception bias is 

 dependent on behavior, weather (e.g. Beaufort sea state), 

 and the observer: active groups are less likely to be missed 

 than resting groups or species whose behavior does not 

 produce pronounced cues (e.g. blows, splashes). 



Availability bias varies by species because of differences 

 in individual dive cycles, group diving behavior, and group- 

 sizes. Long-diving sperm whales and beaked whales will 

 be at the surface for much less time than will many small 

 delphinids, which have much shorter dive cycles. Diving syn- 

 chrony among members of a group also affects availability 

 bias; if dives are a-synchronous, the probability that at least 

 one animal will be at the surface increases with group size. 



Barlow ( 1999) estimated both availability and perception 

 bias for long-diving whales during ship surveys using 25x 

 binoculars in a simulation study and estimated that for 

 dwarf and pygmy sperm whales, Cuvier's beaked whales, 

 and Mesoplodon spp., abundance estimates need to be in- 

 creased 2 to 4 times (i.e.g(0)=0.50 to g( 0=0.25) to account 

 for these biases. Barlow's (1999) estimates of^(O) for per- 

 ception or availability bias (or both) are probably represen- 

 tative of the bias in the southern Atlantic survey because 

 similar ship survey methods were used. However, it may 

 not be valid to apply them directly to our abundance esti- 

 mates because cetacean diving behavior and group sizes 

 may be temporally and geographically specific, and survey 

 conditions and observers may vary among surveys. 



For the strip-transect estimates (Table 2), use of the 

 line-transect strip width (2xl//(0)] from the 25x binocular 

 sightings as the strip width was assumed to be conserva- 

 tive and somewhat negatively biased. The distance from 

 which animals will come to the ship to ride the bow is 

 unknown, and variable, depending on factors such as the 

 animals' previous behavior, number of opportunities for 

 riding bow waves, and the type of ship. If the strip width 

 was too narrow, the strip-transect estimates would overes- 

 timate abundance. 



The geographical bathymetric range of the bottlenose 

 dolphin was not covered during the survey. Because bottle- 

 nose dolphins undertake seasonal movements in the study 

 area, in order to estimate the entire population size, ship 

 survey estimates need to be combined with same-season 

 abundance estimates from coastal waters <10 m and in- 

 shore waters (bays, sounds, and estuaries). 



Distribution 



Water-depth distributions of cetacean species were for the 

 most part similar to those in the Gulf of Mexico (Mullin et al., 

 1994; Davis et al., 1998). Bottlenose dolphins and Atlantic 

 spotted dolphins inhabit the continental shelf and shelf-edge 

 region, whereas most other species have primarily oceanic 

 distributions. The offshore form of the Atlantic spotted dol- 

 phin has not been identified in the northern Gulf of Mexico. 

 The sightings of some species were highly regional (e.g. 

 sperm whales, striped dolphins, Clymene dolphins, pantropi- 

 cal spotted dolphins) were probably heavily influenced by 

 oceanographic features such as the Gulf Stream. Much more 

 survey effort is needed in summer and other seasons before 

 conclusions can be drawn about each species' distribution. 



Acknowledgments 



Many people made significant contributions to the success 

 of this survey including the officers and crew of NOAA ship 

 Relentless and C. Roden, the Field Party Chief The Relent- 

 less was configured for marine mammal surveys through 

 the dedicated efforts of W. Hoggard. The marine mammal 

 observers were C. Brown, C. Burks, C. Gates, W. Hoggard, 

 C. Hubard, T Martinez, K. Maze-Foley, M. Newcomer, S. 

 Swartz, J. Tobias, and K. Touhey. Environmental and ich- 

 thyoplankton data were collected by L. Bero, P. Brown, W. 

 Fambrough, D. Fertl, A. Hamilton, A. Hohn, R. Holmes, E. 

 Keith, E. LaBrecque, J. Litz, and J. Taylor. W. Irvin and T. 

 Pusser were seabird observers. The survey was designed 

 with the help of S. Swartz and the late R. Blaylock. Com- 

 ments by C. Hubard, K. Maze-Foley, and three anonymous 

 reviewers were very helpful in completing the manuscript. 



Literature cited 



Barlow. J. 



IBM. The abundance of cetaceans in California waters. Part 

 I; Ship surveys in summer and fall of 1991. Fish. Bull. 93: 

 1-14. 



