136 
Fishery Bulletin 115(2) 
identified small and large whales; unidentified whales; 
and unidentified cetaceans (Table 1). Sightings of 
unidentified Mesoplodon beaked whales, unidentified 
beaked whales, and rorquals identified as either sei 
or Bryde’s whales were pooled with associated species 
for modeling the detection function (Table 2). Sight¬ 
ings of unidentified small, medium, and large dolphins 
and unidentified dolphins were combined into a single 
category, “unidentified dolphins,” for detection func¬ 
tion and abundance estimation. Likewise, sightings of 
unidentified small and large whales and unidentified 
whales and cetaceans were combined into the category 
“unidentified cetaceans.” 
The detection functions for unidentified rorquals, 
“unidentified dolphins,” and “unidentified cetaceans” 
were estimated separately and without testing for 
the effect of species. The g(0) estimate for unidenti¬ 
fied beaked whales was an average of the estimates for 
Cuvier’s beaked whales and Mesoplodon spp.; the g(0) 
estimate of unidentified rorquals was an average of the 
estimates for fin whales, blue whales, and sei or Bryde’s 
whales; and the g(0) estimate of “unidentified dolphins” 
was an average of the estimates for pantropical spotted, 
striped, rough-toothed, bottlenose, and Risso’s dolphins 
and short-finned pilot whales (Table 3). A g(0) estimate 
was not applied to the “unidentified cetaceans” because 
an appropriate value could not be determined, given the 
broad taxonomic range of this category. 
Results 
Survey sightings 
During the HICEAS in 2010, the systematic and non- 
systematic visual search effort spanned 20,568 km of 
transect lines in Beaufort sea states 0-6 within the 
Hawaiian Islands EEZ. During this effort and while 
off-effort, the observers sighted 379 cetacean groups 
(n=198 during systematic effort, 7Z=101 during nonsys- 
tematic effort, n=80 during off-effort), which include 13 
groups with more than one species present. Accounting 
for these mixed-species groups, the 379 group sightings 
represent 398 sightings of 23 species (17 odontocetes 
and 6 mysticetes) and 13 unidentified species catego¬ 
ries (Table 1). With the exception of the pygmy sperm 
whale (Kogia breviceps) and the extremely rare North 
Pacific right whale (Eubalaena japonica), all cetacean 
species known to occur in the Hawaiian Islands EEZ 
were sighted during the HICEAS in 2010. 
The systematic effort that was relevant to the 
abundance estimation encompassed 16,145 km of tran¬ 
sect lines in Beaufort sea states 0-6 for most ceta¬ 
ceans sighted (Fig. 1), but for pantropical spotted and 
bottlenose dolphins, the effort covered 15,747 km and 
16,100 km, respectively. As with the HICEAS in 2002 
(Barlow, 2006), windy conditions prevailed during the 
HICEAS in 2010, and most (94.5%) of the systematic 
effort occurred in Beaufort sea states 3-6. Adjusting 
for mixed-species groups (n=9), the 198 groups sighted 
on systematic effort correspond to 211 sightings of 20 
species and 11 unidentified species categories (Table 
1; Fig. 2). The 3 species not sighted by the observ¬ 
ers while on systematic effort during the HICEAS in 
2010 were the spinner dolphin, the dwarf sperm whale 
{Kogia sima), and the minke whale {Balaenoptera 
acutorostrata). 
By using the 177 sightings within the respective 
analytical truncation distances (N^st in Table 1), abun¬ 
dance was estimated for 19 cetacean species (15 odon¬ 
tocetes and 4 mysticetes; see the Materials and meth¬ 
ods section for the rationale for excluding humpback 
whales) and for the 11 unidentified species categories, 
although the latter were combined into 6 taxonomic 
categories (as described in the Materials and methods 
section). Of the 48 sightings of unidentified cetaceans 
used in the estimation of abundance, 9 sightings cor¬ 
respond with acoustic detections of dolphin whistles, 
odontocete clicks, or baleen whale calls. These detec¬ 
tions were examined for possible insights into species 
identification. However, this effort did not lead to any 
gains in species identification because of either the 
poor quality of the recordings, the non-specificity of the 
vocalizations, or the confounding presence of an associ¬ 
ated species. 
Detection function 
Of the 6 covariates of interest, only 4 {Beaufort, group 
size, ship, and species) were tested in the 10 models 
of detection function, although only the noncategori- 
cal covariates Beaufort and group size could be tested 
in all cases (Table 2). Insufficient samples sizes by 
cruise number and year prevented testing for the effect 
of these covariates on any of the detection functions. 
Group size and Beaufort most frequently contributed 
to the model-averaged estimates of detection function. 
Specifically, group size was selected in 6 detection func¬ 
tions and Beaufort, in 5 detection functions. 
For the 7 detection functions in which species was 
a consideration, this covariate was tested in 3 cases 
and selected in 2 (Table 2). For the 4 species pools that 
had a limited sample size for testing the effect of spe¬ 
cies, follow-up modeling was performed in 3 cases to 
evaluate the potential for a species effect on the de¬ 
tection function. Specifically, for “species pool 1,” a 
“striped dolphin” and “not striped dolphin” influence 
was examined. For “species pool 3,” the evaluation was 
between “pilot whale” and “not pilot whale” sightings. 
For “species pool 5,” the “other” sighting was excluded 
and a “Cuvier’s beaked whale,” “Mesoplodon spp.,” and 
“unidentified beaked whale” effect was explored. By re¬ 
ducing the number of factor levels, species did enter 1 
of the 4 acceptable models for the “species pool 5” de¬ 
tection function, but this covariate otherwise remained 
unselected for the 3 species pools. Follow-up model¬ 
ing was not undertaken for “species pool 6” because 
there were not enough sightings to evaluate a “sei or 
Bryde’s” and “not sei or Bryde’s” effect. Overall, this 
post-hoc analysis of a species effect produced equivocal 
