138 
Fishery Bulletin 115(2) 
size (10.1 individuals; range: 1.0-20.4 individuals) of 
all available sightings of Longman’s beaked whales 
(71=9) made in the eastern Pacific by the SWFSC be¬ 
fore 2010. Mean ESWs range from 1.61 to 4.42 km, are 
highest for the small delphinids (with the largest mean 
group sizes) and for killer and sperm whales, and are 
lowest for beaked whales (excluding Longman’s beaked 
whales). 
For most species sighted during the HICEAS in 
2010, the proportions of systematic effort in Beaufort 
sea states 0-6 that were used to obtain survey-specific 
estimates of giO) from the values published in Barlow 
(2015) are 0.001, 0.012, 0.042, 0.122, 0.473, 0.304, and 
0.046, respectively. The proportions used for pantropi- 
cal spotted dolphins are 0.001, 0.012, 0.041, 0.124, 
0.474, 0.301, and 0.046, and those used for bottlenose 
dolphins are 0.001, 0.012, 0.042, 0.122, 0.472, 0.303, 
and 0.046. The resulting estimates ofg'(O) (Table 3) are 
substantially lower than those used in the estimation 
of abundance for the HICEAS in 2002 (Barlow, 2006, 
table 2). 
Estimated densities of cetaceans by species and over¬ 
all in the Hawaiian Islands EEZ during the HICEAS 
in 2010 are low (Table 3)—a finding that is consistent 
with results from the HICEAS conducted in 2002 (Bar- 
low, 2006). Estimates of species density do not exceed 
approximately 30 individuals/1000 km^, although more 
than half of the estimates are less than 2 individu¬ 
als/ 1000 km^. Accounting for the estimated density of 
false killer whales (Bradford et ah, 2014, 2015), total 
cetacean density during the HICEAS in 2010 was ap¬ 
proximately 146 individuals/1000 km^. The most abun¬ 
dant species in the Hawaiian Islands EEZ during the 
summer-fall period of 2010 were the rough-toothed, 
striped, pantropical spotted, and Fraser’s dolphins. The 
least abundant species were the blue whale {Balaenop- 
tera musculus), killer whale, and fin whale (B. phy- 
salus). Approximately 4% of the estimated delphinid 
abundance represents unknown species, but more than 
30% of the rorqual abundance and 40% of the beaked 
whale abundance could not be identified to species. The 
estimated abundance of cetaceans with unknown taxo¬ 
nomic status (i.e., “unidentified cetaceans”) is relatively 
low. As expected, given the low number of sightings of 
most species, the CVs for the estimates of density and 
abundance are generally high. 
Discussion 
Although the HICEAS in 2010 was a follow-up survey 
to the HICEAS in 2002, comparisons between the data 
collected and the parameters estimated from the 2 sur¬ 
veys are complicated by several factors. At a basic lev¬ 
el, there is random variation in the sampling process 
(e.g., survey conditions) and in the sighting attributes 
(e.g., group size) of the 2 surveys, and that variation 
can have a pronounced influence on the data and esti¬ 
mates, given the low sighting rates. For example, the 
mean group size of the 1 sighting of Longman’s beaked 
whales made during the HICEAS in 2002 is 17.8 in¬ 
dividuals (Barlow, 2006), compared with the mean of 
59.8 individuals for the 3 sightings during the HICEAS 
in 2010. The single, chance sighting of 100 Longman’s 
beaked whales in 2010 is alone a basis for expecting 
marked differences in the abundance estimates be¬ 
tween the 2 surveys. In addition, although the total 
length of systematic survey effort during the HICEAS 
in 2010 (16,145 km) was similar to that of the HICEAS 
in 2002 (17,050 km), survey coverage within the pelag¬ 
ic portion of the Hawaiian Islands EEZ was somewhat 
greater in 2010 than in 2002 because 3350 km of the 
HICEAS in 2002 was dedicated to an intensive survey 
of the main Hawaiian Islands (Barlow, 2006). This shift 
in survey coverage along with random variation likely 
contributed to differences in the total number and spe¬ 
cies composition of sightings. 
More broadly, there likely was interannual variation 
in oceanographic conditions between the 2 surveys that 
led to differences in the distribution and density of spe¬ 
cies in the study area (Forney et al., 2015). This factor 
becomes particularly important because the Hawaiian 
Islands EEZ is a jurisdictional rather than a biological 
stock boundary, and individuals from many associated 
stocks move into and out of the study area. Therefore, 
apparent differences in species stock density and abun¬ 
dance between the 2 surveys may not represent actual 
changes in the underlying population (or populations), 
but rather indicate a change in the proportion of the 
population within the Hawaiian Islands EEZ. 
Finally, although data collection protocols were con¬ 
sistent and a similar analytical framework was used 
for each survey, differences in the estimation process 
make the resulting estimates difficult to compare. Al¬ 
though sightings from both the HICEAS in 2002 and 
2010 were pooled with sightings from previous surveys 
for modeling detection functions, the pooled sightings 
for the 2010 estimation were limited geographically 
to minimize heterogeneity resulting from geographi¬ 
cal differences in species associations and behavior 
and were further combined with sightings of species 
with similar detection characteristics. Differences in 
the pooled sightings used for modeling the detection 
functions likely partially explain differences in the es¬ 
timates of mean ESW in 2002 and 2010 for many spe¬ 
cies (Barlow, 2006, table 3; Table 3). 
However, the biggest difference in the estimation 
procedure for each survey is the use of the g(0) esti¬ 
mates of Barlow (2015) in the analysis of data from the 
HICEAS in 2010. The present study is the first to apply 
these values to species in the central Pacific, and the 
resulting g(0) estimates (Table 3) are markedly lower 
than those used by Barlow (2006), as well as those 
used in all known previous analyses of line-transect 
surveys of cetaceans. The g(0) estimates in the present 
study reflect the effect of the sighting conditions dur¬ 
ing the HICEAS in 2010, represented by Beaufort sea 
state, and range from being 1.3 times (78.9%) small¬ 
er (i.e., for short-finned pilot whales and Longman’s 
beaked whales) to almost 9 times (11.2%) smaller (i.e.. 
