PERRIN ET AL.: STOMACH CONTENTS OF PORPOISE AND YELLOWFIN TUNA 



and the volume was then measured. Whole un- 

 digested fish were identified by comparison of 

 external characters with published keys or with 

 identified museum specimens and species de- 

 scriptions. Usually though, digestion had pre- 

 cluded such a procedure, so the remains were 

 dissected and skeletal features were compared 

 with a large series of radiographs of determined 

 specimens from the collections of Scripps In- 

 stitution of Oceanography, Fish otoliths extract- 

 ed from the frozen porpoise stomachs allowed 

 identification of completely and nearly complete- 

 ly digested fish. Because the tuna stomachs were 

 initially preserved in unbuffered Formalin, it 

 was not possible to use otoliths from those stom- 

 achs because diagnostic otolith features dis- 

 solve rapidly in an acidic solution. Some rela- 

 tively intact squids (Dosidicus gigas, Symplec- 

 toteuthis oualaniensis, Abraliopsis affinis, and 

 Ouykia sp.) were found and identified in the 

 stomach contents. Beaks from these specimens 

 and from squids taken on the fishing grounds by 

 dipnetting under lights at night were used to 

 identify isolated beaks and beaks associated with 

 otherwise unidentifiable fragments. Published 

 and unpublished keys to cephalopod beaks and 

 beaks from identified squids in the reference col- 

 lection of the Marine Mammal Division, North- 

 west Fisheries Center, were also used in the 

 identifications. Beaks from the ommastrephid 

 squids were relatively easy to identify and sepa- 

 rate from those of squids of other families. Al- 

 though a few of the ommastrephids in the stom- 

 achs were identifiable to species, the data for the 

 Ommastrephidae in the tables below are sub- 

 divided into only three categories: ommastre- 

 phid A (probably Dosidicus gigas), which in- 

 cludes positively identified D. gigas and those 

 fragments (with associated beaks) and isolated 

 beaks most probably belonging to that species 

 on the basis of intrafamilial differences in size 

 and darkening of the beaks; ommastrephid B 

 (probably Sijmplectoteuthis sp.), which includes 

 positively identified S. ouala)iieiisis and those 

 fragments and beaks most probably belonging 

 to that species or another species of Symplec- 

 toteuthis; and u)iide)itified ommastrephid, 

 which includes those fragments and beaks iden- 

 tified only as from ommastrephids. The larger 

 of the two counts of upper and lower beaks was 



interpreted as the minimum number of indi- 

 viduals represented. The only crustacean iden- 

 tified, Enphylax dovii, was present as readily 

 identifiable nearly complete individuals or intact 

 chelipeds. Because many of the porpoise stom- 

 achs contained both the remains (otoliths) of 

 very small fish (myctophids, gonostomatids) 

 and of squid, both of which eat small fish and 

 squid, some of the fish and squid identified were 

 almost certainly present only secondarily. The 

 material was analyzed on a haul-by-haul basis, 

 and the data for each haul are presented and 

 discussed separately. Some effect of the non- 

 selective feeding of tuna, the patchy distribution 

 of prey items, may still be present, however. 

 Since the material analyzed includes some stom- 

 achs that were full of relatively undigested food 

 and some that contained only squid beaks and/ 

 or otoliths from well-digested meals, it is likely 

 that more than one feeding may be represented 

 in the data for some single-net hauls. 



The presentation and analyses of the data are 

 patterned after those in the paper by Pinkas, 

 Oliphant. and Iverson (1971). in which both 

 cephalopod beaks and otoliths were used to 

 identify stomach contents of tunas. The analyses 

 are in terms of volume (for the stomachs ex- 

 amined ashore only), numbers, and percent 

 occurrence. Each of these methods distorts the 

 picture in some way. In the numbers and per- 

 cent occurrence analyses, unimportant small 

 but numerous organisms may be dispropor- 

 tionately evident. The volume analyses are dis- 

 torting because the various organisms are di- 

 gested at different rates and because, as men- 

 tioned above, some of the contents were freshly 

 ingested while others were only the remnants 

 of well-digested meals. Otoliths and squid beaks 

 tend to be retained in the gut. An ideal method 

 would involve extrapolating an estimate of the 

 original volume of each prey item and calcu- 

 lating caloric content, but these api)roaches 

 were beyond the scope of the project. The above 

 caveats notwithstanding, the data do yield 

 considerable information on the relative feeding 

 habits of the three species in mixed aggre- 

 gations. 



The stomachs are considered in three cate- 

 gories: full, with traces, and empty. Full stom- 

 achs are defined as those containing fleshy re- 



1079 



