ANKKNBRANDT: FdOD HABITS OF SKIPJACK TINA 



v/hereN = numerical percentage 

 V = volumetric percentage 

 F = frequency of occurrence percentage. 



2) The mean volumetric ratio measurement 

 (An'RAT) was used to illustrate the biomass impor- 

 tance of prey items without the numeric exaggera- 

 tion implicit in the IRI (John Hedgepeth^). The 

 MVRM was calculated from the volumetric analysis 

 of individual stomachs with each prey item contribu- 

 ting to the total stomach volume. MVRM for each 

 food type is expressed as 



M\'RM = r^ X 100 = mean volumetric percentage 



of prey j to the total volume of 

 n stomachs 



where N = number of stomachs in a given strata 

 y, J = volume of prey type j in stomach i 



V: = Z V: , = total volume of stomach i 



j=i 



'J 



y.- 



'J 



i _ ratio of prey j to the total vol- 



V^ ume of stomach i 



'7 



n = 



'=^^ _ mean volumetric ratio of 



n prey J to the total volume of 



n stomachs. 



Both the IRI and the MVRM, which examine dif- 

 ferent aspects of the diet, were used to evaluate 

 seasonal variations in skipjack tuna food habits. The 

 IRI presents a biased estimate caused by the 

 numerical percentage; the relative importance of 

 small numerous organisms, like euphausiids, is exag- 

 gerated in the IRI because of their high numbers, 

 when actually they may represent the same food 

 value as a few large fish. The M\^RM is an expression 

 of frequency of occurrence and volume without a 

 numeric bias, but does not provide any information 

 on prey abundance. The IRI contains information on 

 the availability of the prey in the environment in 

 terms of numbers, while the MVRM provides an in- 

 dication of its energetic importance to the fish. 



The MVRMs, were stratified by fish length and 



y. B. Hedgepeth, Southwest Fisheries Center La Jotta 

 Laboratorj', National Marine Fisheries Ser\'ice, NOAA, P.O. Box 

 271, La JoUa, CA 92038, pers. commun., April 1982. 



annual quarter (Fig. 2), and tested with the Krus- 

 kal-Wallis nonparametric one-way analysis of 

 variance to evaluate differences in diet with changes 



m size. 



>- 

 O 



z 



UJ 



3 



o 



UJ 



40 50 60 70 80 

 LENGTH (cm) 



Figure 2. - Length-frequency distribution of skipjack tuna from 

 which stomachs were collected. 



381 



