42 
Fishery Bulletin 120(1) 
Scat samples were collected following procedures described 
by Lance et al.” We were careful to ensure that we collected 
each scat in its entirety, we avoided collecting scat where mul- 
tiple scats were likely to be mixed together, and we targeted 
fresh scat (Akmajian et al., 2017) to minimize potential biases 
(Staniland, 2002; Bowen and Iverson, 2013). Our goal was to 
collect 30 scats of Steller sea lions per month and 50 scats 
of California sea lions per season from August 2010 through 
February 2013. We defined seasons as follows: December— 
February as winter, March—May as spring, June-August as 
summer, and September—November as fall. We targeted haul- 
out sites, or sections of these locations, where greater than 
95% of the sea lions counted at the site were of the species 
targeted for scat collections. Scats of California sea lions were 
collected only during the spring, summer, and fall; the diffi- 
culty of landing on East Bodelteh Island in winter meant no 
samples could be collected from the few California sea lions 
that remained at the site. 
Sample processing and prey identification 
Scat samples were washed in a residential-style washing 
machine or through nested sieves by using published pro- 
cedures (Lance et al.°; Orr et al., 2003). Prey hard parts col- 
lected from the scat were dried and stored in glass vials. All 
identifiable hard parts (e.g., bones, otoliths, cartilaginous 
parts, lenses, teeth, and cephalopod beaks) recovered from 
scat were examined by using a dissecting microscope and 
identified to the lowest possible taxon to reduce identifica- 
tion biases (Browne et al., 2002). Prey remains were iden- 
tified by S. Riemer (of the Oregon Department of Fish and 
Wildlife, Marine Mammal Program), who used a reference 
collection of fish and cephalopods from the northeastern 
Pacific Ocean and coastal estuaries (see Riemer et al., 2011). 
Diet analysis 
Diet reconstruction We used split-sample frequency of 
occurrence (SSFO) to reconstruct diets of sea lions because 
the method produces results that are very similar to volu- 
metric estimates of the composition of prey species in the 
diets of predators (Olesiuk et al., 1990) because SSFO can 
be easily incorporated into ecological indices (Krebs, 1999) 
and because it requires only data on presence and absence 
of prey (Laake et al., 2002; Tollit et al., 2007). The SSFO 
approach assumes recovery of remains from all prey con- 
sumed and consumption of all prey at an equal volume 
(Olesiuk et al., 1990). Split-sample frequency of occur- 
rence was calculated by using the following formula: 
* (Cx 1@ 
SSFO, = Yy(On 1) x 100, (1) 
Ss 
where SSFO; = SSFO of taxon i; 
° Lance, M. M.,A. J. Orr, S. D. Riemer, M. J. Weise, and J. L. Laake. 
2001. Pinniped food habits and prey identification techniques 
protocol. Alaska Fish. Sci. Cent., AFSC Process. Rep. 2001-04, 
29 p. [Available from website.] 
Oj; = the absence (0) or presence (1) of taxon i in 
fecal sample k; 
O;, =the total number of all taxa present in 
fecal sample k; and 
s = the total number of fecal samples that con- 
tained identifiable prey. 
We present values of frequency of occurrence (FO) in Sup- 
plementary Table 1 (online only) to allow comparison to past 
studies. The following formula was used to calculate FO: 
FO, 
s O. 
= Dire Pie x 100, (2) 
s 
where FO; = the FO of taxon i. 
We used 2 definitions of prey taxon. For reporting what 
prey sea lions ate, prey taxon was defined as the lowest 
taxon to which a prey item was identified. For calculat- 
ing ecological indices of diet diversity and dietary niche 
overlap, we defined prey taxa by prey family, with the 
exception of flatfish, which were defined by order, and 
cephalopod and amphibian remains, which were defined 
by class, because these groups could not be accurately 
identified to the family level (Sinclair and Zeppelin, 2002). 
We considered taxa that had an SSFO of 5% or greater to 
be common prey. 
Prey diversity and overlap indices We characterized prey 
diversity for the 2 sea lion species by using 2 indices: 
Levin’s niche breadth and the Shannon—Wiener diversity 
index. Levin’s niche breadth is sensitive to changes in 
abundant species, and the Shannon—Wiener index is sen- 
sitive to changes in rare species (Krebs, 1999). We used 
this formula to obtain Levin’s niche breadth values (D): 
= where D is ae: and p; is the SSFO of taxon 1. 
We used this formula to obtain the Shannon—Wiener 
diversity index values (H): 
H= =), x In(p; )). (3) 
We used the Morisita—Horn index (Horn, 1966; Krebs, 
1999) to compare dietary niche overlap of Steller and Cal- 
ifornia sea lions because it is reported to have the least 
bias among indices in comparison of diets when prey pro- 
portions (like SSFO) are used (Smith and Zaret, 1982). 
Morisita—Horn index values (MH) were obtained by using 
the following formula: 
MH = 2) PyPix 
bs > 25 ye 
where p;; = the SSFO of prey taxon i for population j; and 
Di, = the SSFO of prey taxon i for population k. 
(4) 
Dietary niche overlap varies along a scale from 0 to 1, with 
1 indicating a complete overlap and 0 indicating no over- 
lap. We considered a value of 0.65 or greater to indicate 
