34 
Fishery Bulletin 99(1 ) 
125 124 123 122 121 120W 
Figure 3 
On-effort sightings of harbor porpoise recorded in cen- 
tral and northern California during calm (Beaufort 
0-2) sea states. 
survey estimate obtained 1-2 months earlier (A=13,145 
CV=0.39)1 revealed that the 90% confidence interval of dif- 
ferences (CI rf ) did not include zero, indicating that aerial 
and ship abundance estimates were significantly different 
at a = 0. 10 (90% Cl d =-17,275 to -257, P=0.08, Fig. 5, A and 
B). Porpoise abundance for region 2 was estimated at 1041 
(log-normal 95% CI=559-1941; bootstrap CV=0.44, boot- 
strap 95% CI=587-4138), based on 91 km of survey effort 
and 17 porpoise sightings. Our estimate for region 2 was 
significantly different (90% CI f/ =-4084 to -162, P=0.07, 
a=0.10) from an estimate obtained with aerial surveys 1 to 
2 months earlier in the year (A=2861 CV=0.39).l 
In northern California, harbor porpoise were not dis- 
tributed randomly with respect to water depth (Kolmogo- 
rov-Smirnov test, PcO.OOl; Fig. 6A). High proportions of 
porpoise sightings with respect to survey effort were found 
between 20 to 60 m, and fewer porpoise than expected 
were found in waters deeper than 60 m. Because calm 
oajyipcor-N'tiotDWCvi't® 
dooo T-T-d-^ c\i cvj cxi 
Perpendicular distance in km 
Figure 4 
Half-normal model fit to the perpendicular distance data. 
Data were truncated at 1 km, which eliminated 10% of all 
observations. The model fitted the data well, as determined 
by a chi-square goodness-of-fit test (P=0.84). 
sighting conditions often occur in nearshore shallow wa- 
ters where porpoise densities are relatively high (Barlow, 
1988), we also examined relative abundance at depth for 
rough sea states (Beaufort 3-4) to see if the overall depth 
patterns held. The depth distribution of porpoise was sim- 
ilar in rough sea states: more porpoise than expected 
occurred between 20 to 40 m and fewer porpoise than 
expected were seen at depths of 40 to 120 m ( Kolmogorov - 
Smirnov test, P<0.002, Fig. 6B). During our survey, the 
area from shore to the 20-m isobath went largely unsur- 
veyed because of the ship’s draft and the presence of local 
navigational hazards. Owing to an insufficient number of 
sightings within region 2 (n= 25, depth range=20 to 93 m), 
we did not attempt to statistically describe the depth dis- 
tribution of harbor porpoise for that region. 
Discussion 
Abundance of harbor porpoise off northern 
California 
Our estimate of porpoise abundance for northern Cali- 
fornia (5686, log-normal 95% CI=3498-9242) is consider- 
ably lower than estimates for the same area from 1988 to 
1993 aerial surveys (9250, 95% CI=5943-14,397; Barlow 
and Forney, 1994). Barlow and Forney (1994) estimated 
2.05 porpoise/km 2 in northern California from aerial sur- 
veys (corrected for missed groups), whereas we estimated 
1.26 porpoise/km 2 (also corrected for missed groups). From 
1984-85 ship surveys conducted in September, Barlow 
( 1988) estimated 12,700 porpoise (2.09 porpoise/km 2 ) from 
Bodega Head, California to Cape Blanco, Oregon, an area 
approximately 23% larger than our our study area in 
northern California. From aerial survey data collected 1 to 
2 months prior to our ship survey, Forney 1 reported a pre- 
liminary estimate of 13,145 porpoise (CV=0.39, log-nor- 
mal 95% CI=6316-27,357; 2.92 porpoise/km 2 ) for northern 
California. This aerial estimate is significantly different 
from our ship survey estimate, as determined by the con- 
fidence interval of differences test (90% CI rf =— 17,275 to 
-257, P=0.08, a=0.10). Had we used the traditional a level 
