256 



Fishery Bulletin 98(2) 



two overlap indices (Table 4 ). No differences were 

 also found between entangled and nonentangled 

 males in the period 1990-1998 (Table 4). 



When the difference in feeding between sexes 

 was analyzed, no differences were found with 

 the GO, but the SO indicated significant differ- 

 ences in diet (Table 4). Taking into account that 

 the GO analyzes the differences between the uti- 

 lization curves of the groups with reference to a 

 common utilization curve, whereas the SO ana- 

 lyzes one utilization curve with respect to the 

 other one, our data suggest that there are some 

 differences in the diet between sexes. 



Prey size 



No differences were found in the sizes of Ar- 

 gentine hake (^7=13,785.5; ",„„,,,=286; n/,,„„/„= 

 101; P=0.496), Argentine shortfin squid ([/= 

 1,331.5; n,„„,,,=27; n/„„„,^.,= 108; P=0.486), and 

 Argentine anchovy ({7=1,627; /3„,„,„=33; "/„„„,,,= 

 123; P=0.080) consumed by the two sexes. Sev- 

 enty-four percent of the Argentine hake eaten 

 by sea lions were less than 30 cm TL. Argentine 

 shortfin squid consumed by sea lions had a DML 

 greater than 15 cm, whereas Argentine anchovy 

 consumed by sea lions were mostly between 12 

 and 17 cm of TL (Fig. 4). Red octopus consumed 

 by males weighed significantly less than those 

 consumed by females ({7=194. 5; n„,„,,. =21; /!,-„,„., =54; 



•^ ' /?? a / (.'.s ^females ' 



P<0.0001). Patagonian squids consumed by males 

 were larger than those consumed by females ( {7=880; 

 ",„«/.,= 145; /v,„,„,,,= 18; P=0.024), but the range of 

 DMLs of squid consumed by females was gi-eater 

 than that of squid eaten by males. Larger "raneya" 

 was consumed by male sea lions than "raneya" eaten 

 by female sea lions ( {7=2,681; '!,„„/,.,=67; 'ife,„„i,,s=^^^< 

 P<0.0001)(Fig. 5). 



No relationships were found between mean length 

 of prey and predator SL with the pooled sample 

 (r,=0.007; n=37; P=0.964), with males only (r^=0. 004; 

 n = 19; P=0.985) or with females only (r =0.118; n = 18; 

 P=0.641). 



Gastroliths 



Of the stomachs analyzed, 60.4Vf had gastroliths 

 and 87. 97^ had parasites. The parasites found in the 

 stomachs were mostly nematodes. The presence of 

 parasites and gastroliths was independent (Fisher 

 exact test; P=0.999 ). A positive correlation was found 

 between the SL of South American sea lions and GW 

 (r^.=0.572; n=45; P<0.0001) (Fig. 6). 



Gastroliths were found in 56.7'7f of females sea 

 lions and 90.07r of females had parasites. The pres- 



ence of gastroliths and parasites was independent 

 (Fisher exact test; P=0.5645). A positive correlation 

 was found between SL and GW (r^-0.66; n=23; 

 P=0.0006). 



Gastroliths were found in 85.7% of the male sea 

 lion stomachs, and 64.3'^ of males had parasites. 

 The independence between the presence of gastro- 

 liths and parasites could not be rejected (Fisher 

 exact test; P=l ), and a positive correlation was found 

 between SL and GW (/•^,=0.599; «=22; P=0.0032). 



Discussion 



The samples analyzed was collected over a broad 

 range of time and space and were derived from two 

 sampling sources. For that reason several subsam- 

 ples were tested for homogeneity before a sample 

 could be considered to be representative of the diet 

 of South American sea lions. Those subsamples were 

 tested by means of the GO and SO indices because 

 they have associated statistical tests. The GO has 

 only a small bias related to the difference in sample 

 sizes even when the total sample size is relatively 

 small (Smith and Zaret, 1982), and both overlap 

 measures do not change if the resources are divided 



