464 



Fishery Bulletin 88(3), 1990 



59° N 



54° N 



I64°W 



I52°W 



Figure 1 



Location of collection stations in the 

 Gulf of Alaska. Stations represent one 

 transect parallel to the Alaska penin- 

 sula, and two perpendicular offshore 

 transects, with stations 38 and 48 

 located near the shelf break. 



from 15 m off-bottom (to a maximum of 300 m depth) 

 to the surface (Hinckley et al. 1989). Samples from 10 

 stations, representing 2 nearshore-offshire transects 

 (stations 38, 40, and 42 and stations 44, 45, and 48) 

 and one transect parallel to the Alaska Peninsula 

 (stations 21, 30, 32, 42, 44, and 58; see Figure 1), 

 that were collected from 20 to 25 June, were the focus 

 of this study. The total number of fish examined was 

 490. 



Samples were preserved in 5% formalin at sea, and 

 transferred to 70% ethanol prior to their examination. 

 Specimen shrinkage most likely occurred as a result 

 of fixation. 



After the standard length (SL) of each larva was 

 measured, the digestive tract was removed. Contents 

 of the entire digestive tract were evaluated. Gut con- 

 tents were teased out and prey items were identified: 

 invertebrate eggs, copepod nauplii, barnacle nauplii, 

 euphausiid larvae, pteropods, other non-copepod prey, 

 and copepods mduding Pseudocalanus sp. adults (CVI), 

 Pseudocalanus copepodids (CI-CV), Acartia longire- 

 mis, Acartia sp., Calanus marshallae, Calanus sp., 

 Centropages abdominalis, Eucalanus bungii, Metridia 

 sp., Oithcma sp., Oncaea sp., unidentified calanoid cope- 

 podids < 1.5 mm, and unidentified calanoid copepodids 

 >1.5 mm. Copepod eggs comprised over 99% of the 

 "invertebrate eggs" category and Oithona similis com- 

 prised over 99% of the "Oithona sp." category. While 

 the taxonomy of the genus Pseudocalanus has recent- 

 ly been clarified (Frost 1989), no attempt at specific 

 identifications of Pseudocalanus was made in the pres- 

 ent study. For all copepods except Pseudocalanus sp., 

 copepodid (CI-CV) through adult stages were consid- 



ered together. Copepodids were inconsequential in all 

 species except Calanus sp. 



Diet was analyzed in terms of numerical percentage 

 composition (%A''), volumetric percentage composition 

 (%VOL), and percent frequency of occurrence (WoFO). 

 Prey volumes were estimated using prey dimensions 

 (Grover and 011a 1987). Over 8000 prey items were 

 measured for prey volume calculations. The three anal- 

 yses {%N, %VOL, and %F0) were combined to yield 

 a more comprehensive assessment of prey importance, 

 the index of relative importance (IRI = (%N + % VOL ) 

 X %i^0) (Pinkas et al. 1971). Data were pooled across 

 stations for analysis by size class. The four size classes 

 were defined as 10.0-14.9, 15.0-19.9, 20.0-24.9, and 

 25.0-29.9 mm SL. For station-by-station analyses, data 

 were pooled across size classes. 



Diet overlap was calculated using an index of pro- 

 portional similarity (Wallace 1981, Linton et al. 1981, 

 Kohn and Riggs 1982) defined as 



/ " 



PS = 1 



0.5 



I 1^^ 



where PS is the index of proportional similarity or over- 

 lap, Pj., is the proportion of prey category i in the diet 

 at location x, Py, is the proportion of prey category i 

 in the diet at location y. and n is the number of prey 

 categories. Mean percent volume was used as a mea- 

 sure of diet for these calculations (Wallace 1981). Al- 

 though this index has been traditionally used to exam- 

 ine diet overlap between species at one location, I have 

 used it to examine diet overlap between locations for 



