FISHERY BULLETIN: VOL. 86. NO. 3 



18-37 m show no consistent relationship to those 

 at 37-55 m or 55-73 m (Table 3), but on average 

 these appear to be approximately equal. Relative 

 densities at 18-37 m are, however, consistently 

 higher than densities at 73-110 m in all areas 

 (Table 3). A total of 236 km were searched in waters 

 deeper than 110 m and no harbor porpoise were 

 seen. 



Despite high variability in patterns of depth 

 distribution and lack of ship coverage in shallow 

 waters, some generalizations can be made about the 

 depth distribution of harbor porpoise along the west 

 coast. The relative abundance of harbor porpoise ap- 

 pears to be roughly constant from shore to 55 m, 

 to be markedly lower at 73-110 m, and to be very 

 low in waters deeper than 110 m. 



Based on the above relationships, I propose the 

 following preliminary model for the depth distribu- 

 tion of harbor porpoise along the coasts of Califor- 

 nia, Oregon, and Washington: constant abundance 

 from the coast to the 80 m isobath, linearly decreas- 

 ing abundance from the 80-120 m isobaths, and zero 

 abundance in waters deeper than 120 m (Fig. 4a). 

 Because considerable uncertainty exists in this 

 model, I propose two alternative models (Fig. 4b, 

 c). Alternative models b and c are less likely than 

 the primary model given because both conflict with 

 some of the available data. The alternative models 

 do, however, encompass the likely range of relative 

 density values and provide a means to evaluate the 



sensitivity of the abundance estimate to different 

 models of depth distribution. 



Fraction of Missed Animals 



The experiment on survey 4 indicates that some 

 trackline groups were seen by 1 group of observers 

 and were missed by the other. A total of 103 sight- 

 ings was made by both teams, 33 of which were 

 estimated to be within 100 m perpendicular distance 

 from the transect line. Of the 103 total sightings, 

 85 were detected only by the 5 principal observers, 

 6 were detected only by the 3 monitor observers, 

 and 12 were detected by both teams. Of the 33 

 trackline sightings, 20 were detected only by the 

 principal observers, 3 were detected only by the 

 monitor observers, and 10 were detected by both 

 teams. The Petersen estimate of the fraction of 

 trackline porpoise seen by the primary team of 5 

 observers is thus 0.780 (SE = 0.117, 95% C.L. = 

 0.45-0.95). This indicates that approximately 22% 

 of trackline sightings are missed by the principal 

 teams of 5 observers. 



Porpoise Abundance 



Estimates of porpoise abundance in each of the 

 eight geographic strata are given in Table 4 for the 

 primary model of offshore distribution. Independent 

 estimates are given for survey 1 and for survey 3 



Table 3. — Relative harbor porpoise abundance observed within the specified depth 

 ranges at a variety of study sites. Relative abundance is measured as number 

 of porpoise sightings made per kilometer. Numbers in parentheses indicate 

 kilometers surveyed. 



'Data taken from LaBarr and Ainley (see text footnote 8) assuming an average survey speed 

 of 9 knots. 



^Data taken from Szczepaniak (see text footnote 9) assuming an average survey speed of 

 9 knots. 



426 



