926 



Fishery Bulletin 101(4) 



increasing^'. In general, the arithmetic mean of group-size 

 may be an overestimate of the true mean group-size and 

 could lead to positively-biased abundance estimates. There- 

 fore, a regression of group-size by j' was used to estimate an 

 "expected mean group-size" (program DISTANCE) and it 

 was used if the regression was significant (P<0. 15). VarCS^^) 

 was the analytical variance for mean group-sizes based on 

 arithmetic means or was estimated as in Buckland et al. 

 (2001:74) for expected mean group-sizes. 



Strip-transect estimates 



One requirement for unbiased line-transect estimates of 

 abundance is that the cetacean group should not move in 

 response to the ship before it is sighted (Buckland et al., 

 2001). If cetaceans are not sighted before they respond 

 to the ship, in cases of attraction to the ship, /TO) and 

 abundance will be overestimated. During previous U.S. 

 GOM surveys, groups of T. truncatus or S. frontalis were 

 consistently attracted to ride the bow waves as the ship 

 approached (Wiirsig et al., 1998). Therefore, the abundance 

 and variance of groups sighted by naked eye (Ng) were 

 estimated by 



N., 



2-L-w, 



and 



var(/V„J = N=.^ 



var(/i^,J ^ ^ar(-^5,.;) 



(3) 



(4) 



where w^ = l//",(0) which was treated as a constant, i.e. strip 

 width, w^ , was equal to the line-transect effec- 

 tive strip half-width [l//](0)] with vardu^) = 0. 



For each region, species total abundance (Nj.^ ) was the 

 line-transect and strip-transect estimates added, iV^.^ = 

 Nj^, j + Ns,,j. Total U.S. GOM OCS abundance for each 

 species was Nj,^ = 27V j., ^. The coefficient of variation (CV) 

 for each abundance was estimated as CViN) = [vartAOl'^^W 

 and the CV for each summed abundance as 



CV(/V,„„) 



■(I 



CV'{N)N 



r/1 



N. 



(5) 



Results 



Abundance estimates were based on 2196 km of effort and 

 140 sightings (Figs. 1 and 2). For east and west regions, 

 there was 816 km of effort and 73 sightings, and 1380 km of 

 effort and 67 sightings, respectively (Tables 1 and 2). Only 

 three cetacean species were encountered. Groups of J! trun- 

 catus (30 east region, 45 west region) and S. frontalis (34 

 east, 12 west) were the most frequently encountered (Fig. 

 2, Table 2) and S. bredanensis groups ( 1 east, 2 west) were 

 also sighted. Tiirsiops truncatus and S. frontalis were esti- 

 mated to have flO) of 0.6238/km (CV=0.12) and 0.4101/km 

 (CV=0.11), and an effective strip half-width of 1603 and 



2438 m, respectively (Figs. 3 and 4). Steno bredanensis and 

 T. truncatus+S. fron talis abundances were based on an /! ) = 

 0.6059/km(CV=0.11) and an effective strip half-width of 

 1650 m. 



Mean group-sizes (from 25x binocular sightings) of T. 

 truncatus for east (9.8, 0.25) and west (10.0, 0.18) regions 

 were similar (Table 2), and had an overall range of 1-68 

 animals. The mean group size of S. frontalis was larger 

 in the east (24.3, 0.19) than the west (15.6, 0.21) with an 

 overall range of 1-267 animals. Group-sizes of S. bredanen- 

 sis were 8, 11, and 20 animals. The east mean group-size 

 for both T. truncatus and S. frontalis is the size-biased 

 or expected mean group-size because the expected mean 

 was significantly smaller that the arithmetic mean, 10.9 

 (P=0.14) and 31.9 (P=0.08), respectively 



The most abundant species (number of individuals; CV) 

 found in U.S. GOM OCS waters was S. frontalis (30,772; 

 0.27); the vast majority (91%) occurring in the east 

 (27,997; 0.29). The density of S. frontalis was about eight 

 times greater in the east compared to the west (20.1 and 

 2.6 dolphins/100 km^, respectively). The abundance of T. 

 truncatus was 25,320 (0.26); there was greater abundance 

 in the east (15,198; 0.34) than m the west (10,122; 0.29) 

 but with similar densities (10.9 and 9.5 dolphins/100 km^, 

 respectively). The total OCS abundance of S. bredanensis 

 was 1238 (0.65), and that of 7: truncatus+S. frontalis, 1868 

 (0.37). 



Discussion 



Both T. truncatus and S. frontalis occur in northern GOM 

 waters outside the OCS (i.e. waters <20 m or >200 m). 

 About 23,000 T. truncatus inhabit inshore and coastal 

 waters (<20 m) (Waring et al., 2001) and nearly 3000 occur 

 in oceanic waters (Mullin and Fulling-). Both the "coastal" 

 and "offshore" ecotypes of T! truncatus (Hersh and Duffield, 

 1990) occur in the northern GOM (LeDuc and Curry 1998). 

 How these ecotypes are distributed in the northern GOM 

 and western North Atlantic is being investigated from skin 

 biopsy samples collected, in part, during the 1998-2001 

 OCS surveys. Using mitochondrial DNA, obtained from 

 biopsy samples collected during a U.S. Atlantic ship survey, 

 Torres et al. (2003) reported no offshore form was sampled 

 within 6 km of shore and no coastal from was sampled 

 beyond 39 km from shore or in waters >34 m deep. Forty- 

 seven percent (35/75) of the GOM OCS T. truncatus groups 

 were in waters >34 m deep. 



Ship surveys of northern GOM waters indicate that very 

 few S. frontalis (<500 animals) occur in oceanic waters, and 

 those that do are usually found close to the shelf edge in 

 waters <500 m deep (Davis et al, 1998; Mullin and Full- 

 ing2). The smaller "offshore" or "Gulfstream" S. frontalis 

 that occurs in parts of the oceanic Atlantic (Perrin, 2002) 

 has not been recorded for the northern GOM. During the 

 1998-2001 surveys, S. frontalis was sighted in waters <20 

 m deep. However, because sampling was not perpendicular 

 to bathymetry, abundance estimates were not calculated. 

 This species is not known to occur in U.S. GOM inshore 

 waters (Mullin and Hansen, 1999). 



