NOTE Zavala-Gonzalez and Mellink: Entanglement of California sea lions in fishing gear 
181 
Figure 1 
Central-northern Gulf of California, showing the study localities. 
young animals. We excluded dead animals because 
time since death could not be assessed. Pups spend 
most of their time on land and have virtually no in- 
teraction with fishing gear during the survey peri- 
ods. When we made more than one survey per year, 
we considered the highest rate of entanglement as 
the best estimator. 
We used the “differences-between-proportions” test 
(Zar, 1974) to compare the proportions of age and 
sex classes between the entangled animals and the 
population. Differences between sea lion colonies and 
years were compared by using a two-factor Analysis 
of Variance (ANOVA) and Newman-Keuls (N-K) tests 
(Zar, 1974) on the basis of the number of entangled 
animals (adjusted by a square-root transformation) 
and entanglement rates (adjusted by an arcsine 
transformation). Variation in the number of en- 
tangled animals and in the average transformed en- 
tanglement rates over time, as well as the relation- 
ship between number of entangled animals and 
colony size, were analyzed with simple linear regres- 
sions (Zar, 1974). 
In the two cases where we lacked data, we used 
the eight colonies, for which we had complete infor- 
mation, to calculate the relationship between the year 
with missing data (1991 or 1992) and the average of 
the remaining four years, and then used the average 
of the problem colony (Los Cantiles or San Pedro 
Martir) to estimate the missing value. This was done 
in order to perform the statistical tests on a stan- 
dard basis. 
Types of fishing gear involved were recorded from 
dead entangled animals. This information was com- 
pleted with records about the size and characteris- 
tics of the fishing gear used by fishermen working in 
the islands. Similarly, fishing gear debris found on 
sea lion rookeries and showing evident signs of hav- 
ing been associated with a sea lion (bites, sea lion 
fat,) was noted. We also interviewed local fishermen 
about their problems with sea lions. 
Results and discussion 
Sea lion entanglement 
During the study, we counted 237 entangled animals 
(Table 1), 207 of which could be assigned to a par- 
ticular sex and age class: 46.4% were young animals 
(1-3 yr), 41.5% females, 7.2% subadult males, and 
4.8% adult males. The percentage of young animals 
among the entangled animals was statistically higher 
than their proportion in the censused population 
(25.8%, Z=6.70, P=7.13xl0 -11 ), whereas that of fe- 
males was lower (60.5%, Z--5.53, P=0. 913x10 8 ). 
This is probably a result of young animals being more 
curious, less experienced, and weaker, in addition to 
foraging closer to the surface (senior author, personal 
observation [1994-95]). In other species (northern 
fur seals, Callorhinus ursinus ) high rates of juvenile 
mortality might have caused a decrease in the popu- 
lation (Trites, 1992). Our data are not sufficient to 
establish or discard any such links. Percentages of 
entanglement of subadult and adult males were not 
different from percentages of proportion of subadult 
and adult males in the population (6.3%, Z=0.78, 
P=0.294 and 7.4%, Z=1.5, P=0.13, respectively). 
Between 1991 and 1995, the number of recorded 
entangled sea lions for all 10 rookeries combined 
varied from 34 (±2.27, 95% Cl) (adjusted to 36) to 
72, with a low of 24 (±4.02, 95% Cl) (adjusted to 27) 
in 1992 (Table 1). Regression analysis between num- 
ber of entangled animals and year was significant 
(n= 5, r 2 =0.79, F=11.05, P=0.045). Significant differ- 
ences were found only between 1992 and 1995: the 
other years were not different from one other (F=3.09, 
P=0.027). 
In 1992, there were not only fewer entangled ani- 
mals, but also there were fewer sites with entangled 
animals. This year saw the worst recent fishing sea- 
son in the central Gulf according to information at 
the Bahia de Los Angeles fishing office, and this find- 
ing is likely a reflection of the prevailing El Nino 
Southern Oscillation conditions, which had strong 
