Uchikawa et al.: Diet of Berryteuthis anonychus in the northeast Pacific during spring 



735 



and an asymmetrically swollen genital segment. The gen- 

 eralized form included calanoid copepods of the Calanus 

 type that did not share any of these characters. 



A stomach-contents index (SCI, %) was calculated 

 as SCI=(wet mass of total stomach contents/wet body 

 mass)xl00. For each prey taxon, the percentages by 

 number (N) and wet mass (WM) of the total prey, and 

 the percentage frequency of occurrence (F) were de- 

 termined. An index of relative importance (IRI) was 

 calculated for each prey taxon as IRf = F t x lN t + WM t ) 

 (Pinkas et al., 1971), where i denotes the taxon. The IRI 

 for each major group of prey taxa was then standard- 

 ized to '/dRI (Cortes. 1997): 



n 



%IRI, = 100 x IRI, I £ IRI t , 



where /; is the total number of groups collected. 



Copepod mandible size is directly related to the cara- 

 pace length of several calanoid copepods in the North 

 Atlantic (Karlson and Bamstedt, 1994); therefore man- 

 dible width was used as an indicator of relative prey 

 size to compare copepod prey size with squid mantle 

 size. A total of 87 mandibles were measured from the 

 stomachs of 10 squid measuring 29-102 mm ML. 



Results 



Of the 338 stomachs examined, 267 (79%) contained 

 prey, and 220 (65% ) contained identifiable prey (Table 1). 

 Individual SCI values ranged from 0% to 8.0% (station 

 mean = 1.0%). SCI values varied significantly among 

 sampling times (Kruskal-Wallis test, P<0.001), and the 

 two highest SCI values occurred in the afternoon and 

 just after sunset (Fig. 3). 



The diet of B. anonychus comprised seven major 

 prey groups and was dominated by copepods (A?=70%, 



WM=85%, F=74%, %IRI=87%) and chaetognaths 

 (N=24%, WM=11%, F=48%, <7dRI=129c) (Table 2). The 

 five other prey groups (amphipods, euphausiids, ostra- 

 cods, unidentified fish, and unidentified gelatinous prey) 

 each had a 9cIRI value <1%. 



Copepod prey comprised four genera, and 86% by 

 number of the copepods were from the genus Neocala- 

 nus. Neocalanus cristatus was the most abundant prey 

 taxa. composing 50% by mass and 35% by number of 

 the total diet. The three Neocalanus taxa (Neocalanus 

 spp., N. plumchrus+flemingeri, and N. cristatus) com- 

 posed 85% by mass and 68% by number of the diet. 

 Neocalanus cristatus was identified based on the pres- 

 ence of the head crest, which develops at the C5 copepo- 

 dite stage (Brodsky, 1950). Thus, this taxon comprises 

 only the C5 and C6 stages, and possible members of 

 the Neocalanus spp. taxon include N. plumchrus, N. 

 flemingeri, and earlier stages (C1-C4) of N. cristatus. 

 Squid >60 mm ML fed mainly on Neocalanus crista- 

 tus (2V=39%, WM=53%, F=50%) and Neocalanus spp. 

 (iV=29%, WM= 31%. F=40%), whereas those <60 mm 

 ML fed mainly on Neocalanus spp. (AT=43%, WM=53%, 

 P=29%) and Neocalanus plumchrus+flemingeri (N=8%, 

 WM=10%, F=14%), and consumed few C5-C6 Neocala- 

 nus cristatus (N=4%, WM=4%, F=6%). The mandible 

 size of copepod prey showed a clear positive relationship 

 with ML (Fig. 4), indicating that the squid fed on larger 

 copepods as the squid grew. Taxa from other copepod 

 genera (i.e., Candacia, Metridia, and Pleuromamma) 

 composed 0.5% of the total prey number and 0.1% of 

 the total wet mass (Table 2). 



Parasagitta elegans, the only identified chaetognath, 

 occurred in more stomachs (47%) than any other prey 

 taxon and in 58% of the stomachs from squid >60 mm 



