Scordino et al.: Dietary niche overlap and prey consumption for Eumetopias jubatus and Zalophus californianus 43 
significant niche overlap (see Aurioles-Gamboa and Cama- 
cho-Rios, 2007; Orr et al., 2011). 
To accompany the prey diversity and niche overlap 
indices, we obtained corresponding 95% nonparametric 
bias-corrected and accelerated bootstrap confidence inter- 
vals using 5000 bootstrap samples. The bias-corrected 
and accelerated bootstrap confidence interval takes into 
account potential bias and skewness in a bootstrap sam- 
pling distribution (DiCiccio and Efron, 1996). The pack- 
age boot (vers. 1.3-20; Davison and Hinkley, 1997; Canty 
and Ripley, 2017) was used in statistical software R (vers. 
3.6.2; R Core Team, 2019) to calculate each bootstrap sam- 
ple, bootstrap sampling distribution, and bootstrap con- 
fidence interval. Each bootstrap sample was taken with 
replacement within a given sampling strata. 
Prey consumption model 
We developed models to estimate average seasonal and 
annual prey consumption for Steller and California sea 
lions by using published data sources and data collected 
during this study for sea lions in northwest Washington 
for the period from 2010 through 2013. 
Simply stated, the model multiplies an estimate of prey 
consumed per individual per day by the total number of 
sea lions present in the study area and extrapolates out, by 
using the number of days in the season, to estimate total 
prey consumed each season (see Equation 5). We estimated 
consumption for spring (March—May), summer (June— 
August), fall (September—November), and winter (Decem- 
ber—February) for both California and Steller sea lions. 
Added together, the seasonal estimates provide a yearly 
estimate of total consumption for each sea lion species. 
To determine annual prey consumption by California 
and Steller sea lions in northwest Washington during 
2010-2013, we used the following equation: 
Nin ag in an 3 Cag 8 CES Bg 3 Damn Sif) / MOOD, () 
where x = season; 
y = the species of sea lion (California or Steller sea 
lion); 
z = the demographic group (adult male, adult female, 
juvenile male, and juvenile female for Steller sea 
lions and adult males for California sea lions); 
W,, = the estimated average body weight in kilograms 
of sea lions of species y and demographic group z. 
cy, = a conditional parameter for the percentage of 
body weight that the average sea lion of species 
y and demographic group z eats per day. 
d,, = the number of days in season x; 
Nyy = the average count of age-1+ sea lions hauled out 
in the survey area during season «x for species y; 
Pxyz = the proportion of sea lions counted in season x, 
of species y and demographic group z.; and 
fx = the correction factor for converting the haul-out 
count to the total abundance of sea lions in the 
environment of northwest Washington during 
season x for species y. 
California sea lion average body weight was informed 
by the average body weight of male California sea lions 
captured at Astoria, Oregon (Wright et al., 2010), and 
Ballard, Washington (Gearin et al., 2017). The w,, 
for Steller sea lions was informed by published body 
weight estimates for demographic group z (Winship 
et al., 2001). 
The conditional parameter c,, was informed by pub- 
lished bioenergetics modeling estimates of Winship et al. 
(2006) for male California sea lions and by demographic 
group z for Steller sea lions. The bioenergetics models of 
Winship et al. (2006), the methods of which are described 
in greater detail in Winship et al. (2002), incorporate ener- 
getic costs for lactation and gestation. 
The average count of age-1+ sea lions was informed by 
our surveys as previously described. 
We assumed that California sea lions composed a sin- 
gle male demographic group. The p, ,, for adult female 
and adult male Steller sea lions were informed directly 
by the haul-out demographic counts. We were unable 
to differentiate juvenile Steller sea lions by sex during 
counts and had to calculate the expected proportion of 
juveniles that were male or female on the basis of sex- 
based survival estimates through age 5 for Steller sea 
lions branded in Northern California and southern Ore- 
gon (Wright et al., 2017), assuming an equal sex ratio 
at birth. 
We used a correction factor developed by Lowry and 
Forney (2005) for counts of California sea lions from 
aerial surveys at haul-out sites in Northern California. 
For Steller sea lions, we used the reciprocal of the 
proportion of age-1+ sea lions hauled out in southern 
British Columbia (Olesiuk*). Olesiuk* reported that 
36% (coefficient of variation [CV]=2.1%) of age-1+ sea 
lions were hauled out in winter during the time frame of 
1000-1800, when our surveys were typically conducted, 
and they reported that 67.4% (CV=5.6%) of age-1+ sea 
lions were hauled out in summer. We applied the win- 
ter correction factor for Steller sea lions to the spring 
and fall because Olesiuk* found no significant differ- 
ence in the proportion of time that satellite-tagged sea 
lions hauled out in the spring and winter and because 
Whitlock et al. (2020) found much lower haul-out atten- 
dance in the spring than in the summer. The correction 
factors calculated from data presented by Olesiuk* fall 
within the range and seasonality of the proportion of 
sea lions observed hauled out by Whitlock et al. (2020). 
No correction factors were applied for sea lions missed 
by vessel- or land-based surveys that would have been 
visible and counted during aerial surveys (Westlake 
et al., 1997) because no correction values were available 
in the literature. Likewise, no correction factors were 
applied to estimate the number of sea lions that were 
using haul-out sites off southern Vancouver Island, 
British Columbia, and that would forage in the same 
marine area as the sea lions that were using haul-out 
sites in northwest Washington. 
We divided our total consumption estimate by 1000 to 
convert our estimate from kilograms to metric tons. 
