Jacobson et al.: Use of parasites to clarify migration of Sardinops sagax 
199 
Because Lo et al. (2011) reported different migration 
rates (estimated percentages of the total biomass) of the 
northern subpopulation of Pacific sardine on the basis of 
size, we initially analyzed non-migrant Pacific sardine 
(<200 mm SL) and different size categories of potential 
migrants separately. Pacific sardine >200 mm SL (poten¬ 
tial migrants) were split into 3 size categories: 200-209, 
210-219, and >220 mm SL. 
We used multivariate methods to explore variation in 
the parasite community data. Multivariate analyses of 
parasite communities were performed by using the mul¬ 
tivariate statistics package PRIMER, vers. 7 (PRIMER-e 
Ltd., Auckland, New Zealand) (Clarke and Gorley, 2015) 
that includes the PERMANOVA+ module (Anderson et al., 
2008). Only infected fish were included in multivariate 
analyses. Sample sizes of infected Pacific sardine per size 
category per region are provided in Table 2. Some individ¬ 
ual fish harbored few parasites; therefore, parasite inten¬ 
sity data were averaged among 3-5 fish with the greatest 
similarity in length. Parasite intensity data were square- 
root transformed to reduce skewness due to any dominant 
parasite species and to stabilize the variance (Clarke et al., 
2014a). Unidentifiable parasites (41 of the 2451 parasites 
recovered, <2% of the total) were not included in the multi¬ 
variate analysis. 
We constructed resemblance matrices based upon 
Bray-Curtis similarity with a dummy parasite value of 1 
included to reduce the variability of some of the sparsely 
infected samples (Clarke et al., 2006). We then used these 
matrices to construct nonmetric multidimensional scaling 
(MDS) ordinations on regionxfish size centroids to graph¬ 
ically explore spatial patterns of differences among the 
parasite communities of fish groups (Clarke et al., 2014a). 
To quantitatively evaluate variation in parasite commu¬ 
nities, multivariate parasite data were modeled for main 
effects by using a 3-way permutational multivariate anal¬ 
ysis of variance (PERMANOVA) based on Bray-Curtis 
similarities (Bray and Curtis, 1957). The PERMANOVA 
test applies a nonparametric discriminative method based 
on permutation tests that do not rely on assumptions (e.g., 
normality, equal variance) commonly too stringent for most 
ecological data sets (Anderson, 2001). Main factors in the 
model include region, fish size category, and haul nested 
in region as a random factor. Year was included as a ran¬ 
dom factor in the initial model, but it was not a significant 
factor and was removed. The models were run for 9999 
Table 1 
General information on Pacific sardine (Sardinops sagax) classified as either non-migratory (<200 mm in standard length [SL]) 
or migratory (>200 mm SL). Fish were caught by using a modified rope trawl or purse seine in 5 regions of the California 
Current System from 2005 through 2008 and analyzed for parasites. H=total number of hauls; n=number of fish caught; and 
SE=standard error. 
Mean SL 
<200 mm >200 mm 
Region 
H 
n 
Years 
(SE) 
Min. 
n 
Years 
(SE) 
Max. 
Vancouver Island, Canada 
6 
43 
05-07 
175.8 (2.68) 
148.1 
176 
05-07 
223.1 (1.14) 
285.7 
Washington and Oregon 
19 
113 
05-07 
184.6(1.55) 
100.6 
419 
05-08 
219.4 (0.83) 
279.6 
Northern California 
10 
142 
06-07 
191.8 (0.78) 
120.2 
139 
06-07 
208.2 (0.82) 
260.4 
Central California 
5 
152 
05-07 
175.4(1.23) 
106.8 
36 
06-07 
211.3 (1.88) 
267.2 
Southern California 
4 
112 
06-07 
178.3 (0.93) 
157.2 
56 
06-08 
220.8(1.53) 
257.9 
Table 2 
Number of Pacific sardine (Sardinops sagax) caught from 2005 through 2008 in 5 regions of the Cali¬ 
fornia Current System and infected with at least 1 parasite to be included in multivariate community 
analyses. SI -standard length. 
Size category 
(mm SL) 
British 
Columbia 
Washington and 
Oregon 
Northern 
California 
Central 
California 
Southern 
California 
Total 
<200 
29 
40 
87 
79 
87 
322 
200-209 
20 
55 
59 
7 
7 
148 
210-219 
40 
72 
13 
11 
9 
145 
>220 
78 
101 
6 
2 
21 
208 
Total 
167 
268 
165 
99 
124 
823 
