NOTE DeMaster et al.: Effect of sea state on estimates of abundance of Delphinapterus leucas 
199 
Table 2 
Summary of beluga whale densities (animals seen/km 2 ) from Norton Sound, Alaska, by Beaufort (BF) sea state number. Numbers 
in parentheses represent 95% confidence intervals. 
BF sea state 
1993 
1994 
1995 
1 
0.217 (0.115-0.409) 
0.107 (0.048-0.243) 
0.294 (0.122-0.710) 
2 
0.111 (0.060-0.207) 
0.052(0.032-0.087) 
0.046 (0.015-0.142) 
3 
0.039(0.022-0.069) 
0.138 (0.075-0.256) 
0.111 (0.043-0.288) 
4 
0.009 (0.003-0.031) 
0.075 (0.015-0.367) 
insufficient survey effort 
Table 3 
Summary of sighting statistics for beluga whales in Norton Sound, Alaska, by Beaufort (BF) sea state, for 1993-95 data combined. 
Densities were estimated as weighted average, where weights were proportional to survey effort. 
BF sea state 
No. of observations 
Survey effort ( km ) 
Average density (SE) (animals seen/km 2 ) 
1 
653 
2343 
0.206 (0.051) 
2 
416 
3916 
0.071 (0.014) 
3 
397 
3618 
0.069(0.013) 
4 
19 
490 
0.053 (0.026) 
The line-transect analysis was stratified by BF sea state. 
Using Eberhardt’s ( 1968) multiple comparison test, we did 
not reject the null hypothesis that sea state specific den- 
sity was independent of survey year; therefore, sighting 
data by BF sea state were pooled across years, where aver- 
age density for each sea state was calculated as a weight- 
ed average proportional to distance searched. Differences 
in average density by BF sea state pooled across years 
and in the effective strip width by BF sea state in a given 
year were also tested using Eberhard’s multiple compar- 
ison test, whereas differences in the average encounter 
rate by BF sea state were tested by using a goodness-of- 
fit test based on the number of sightings and survey effort 
for each BF sea state. All statistical tests were performed 
with the type-I error set at 0.05. The variance of the ratio 
of the estimated average density in BF sea state 1 to the 
weighted average density in BF sea states 2, 3, and 4 was 
estimated with the delta method (Seber, 1973). Search ef- 
fort in BF sea state 0 and BF sea state 5 was conducted 
only in 1995 and was relatively small (e.g. 264 km of effort 
in BF sea state 0 and 50 km of effort in BF sea state 5); 
therefore, sightings data for these sea states were not used 
in our analysis. 
Density estimates reported in our note have not been 
corrected for either the period of time that animals were 
underwater (and therefore not observable), or the num- 
ber of animals at the surface that were missed. Further, 
it was necessary to assume that the actual density of be- 
luga whales was independent of sea state and was rela- 
tively constant between years. Given the distribution of 
sightings observed during the three years of survey effort, 
this assumption seemed reasonable, except for the north- 
ern portion of Norton Sound, where beluga whales were 
not observed (Fig. 1). The proportion of total survey effort 
conducted in this area was relatively small (8%); whereas 
the percent of survey effort conducted in BF sea-state- 1 
conditions in this area was approximately three times the 
percent of survey effort in the overall survey (i.e. 63% vs. 
23%). The effect of this heterogeneity in the distribution 
of sightings in relation to BF sea state was assumed to be 
negligible, although it was recognized that the bias associ- 
ated with this factor would produce negatively biased den- 
sity estimates in BF sea state 1 in contrast to other sea 
states. 
Results and discussion 
The yearly density estimates for beluga whales in June of 
1993, 1994, and 1995 by sea-state category are presented 
in Table 2. In two of the three years, the highest annual 
density estimate always occurred in BF sea state 1. For 
the pooled data for all three years (Table 3), the encoun- 
ter rate (i.e. number of sightings per km of survey effort) 
was not random with respect to sea state (chi square=403, 
PcO.OOl); the largest chi square value was associated with 
the number of sightings in BF sea state 1. In addition, 
the estimated density for BF sea state 1 was significantly 
greater than the estimated density for BF sea states 2, 3, 
or 4. There were no significant differences between esti- 
mated density for BF sea states 2, 3, and 4, although the 
estimated density for BF sea state 4 was approximately 
50% of the estimated density in BF sea states 2 and 3. 
These data are consistent with the hypothesis that the 
probability of sighting beluga whales in BF sea state 1 is 
significantly greater than that for sighting beluga whales 
