YANG and LIVINGSTON; FOOD HABITS OF GREENLAND HALIBUT 



Stomachs were analyzed individually in the labor- 

 atory. Prey items were identified to the lowest possi- 

 ble taxonomic level and counted. Wet weights of the 

 prey items were recorded to the nearest milligram 

 after blotting with paper towels. The fork lengths 

 of prey fish were also measured. 



Diet Description 



Since the depth distribution of Greenland halibut 

 in this study was broad (from 62 m to 891 m), stom- 

 ach content data were first subjectively divided by 

 100 m depth groups. For each 100 m depth class, 

 percent frequency of occurrence (%F0) of prey 

 items, percentage of total stomach content weight 

 (%W) by prey t}T)e, and the percentage of total prey 

 number (%A^) by prey type were calculated by using 

 ECO/INDEX, a computer program for calculating 

 feeding ecology indices (Vodopovich and Hoover 

 1981). Based on similarities of major prey items 

 (using percent by weight), stomach content data 

 were combined into four depth groups for analysis: 

 Depth 1 (<200 m). Depth 2 (200-399 m), Depth 3 

 (400-599 m), and Depth 4 (^600 m). 



Within each of the four depth groups, data were 

 subjectively divided by fish length into 10 cm size 

 groups. By comparing percent by weight of the 

 major prey categories (e.g.. gadids, squids) for each 

 size group, the fish within each of the four depth 

 groups were finally lumped into five size groups: <20 

 cm, 20-29 cm, 30-49 cm, 50-69 cm, and >70 cm. 

 Seasonal breakdowns of stomach contents by depth 

 and predator si^^e group could not be performed due 

 to insufficient sample sizes. 



Diel Feeding Pattern 



Because of the small sample size of fish <30 cm 

 long, only data from three size groups (30-49 cm, 

 50-69 cm, >70 cm) were used for diel feeding anal- 

 ysis. For each size group, the stomach content 

 weights as percent of body weight were calculated 

 for each 3-h period of the 24-h day. Any possible 

 seasonal variations of the diel feeding pattern could 

 not be analyzed because of insufficient seasonal 

 samples. 



Daily Ration 



In this study, daily ration was calculated using 

 Elliott and Persson's (1978) model. The basic 

 assumption of this model are that the rate of gastric 

 evacuation (R) is exponential and temperature 



dependent. If stomach samples are taken at fixed 

 intervals of t hours, the mean stomach content 

 weight as a percentage of fish weight (S,) in each 

 interval (i) is calculated for a total of m intervals 

 over the 24-h period. According to Elliott and 

 Persson (1978), the daily ration (D.R.) in terms of 

 percentage of body weight is therefore given by 



D.R. = 



Rt 



1 - expi-Rt) ' = 1 



24 SR 



J. 8,(1 - expi-Rt)) 



(2) 



where S = J. S,/m. Elliott (1972) found the gen- 

 eral relationship between R and temperature (T) 

 was exponential: 



R = ae 



hT 



(3) 



Based on data presented in the literature for the 

 normal temperature range of both freshwater and 

 marine fishes, Durbin et al. (1983) concluded that 

 the slope (b) is fairly constant for different prey 

 types and fish species (mean = 0.115), while the 

 intercept (a) changes with prey type and can be 

 estimated from gastric evacuation rate experiments. 

 Since there were no gastric evacuation rate data 

 available for Greenland halibut, results of gastric 

 evacuation experiments on walleye pollock, Thera- 

 gra chalcogramma, feeding on juvenile pollock and 

 squid were used (Dwyer et al. 1987). Although wall- 

 eye pollock is taxonomically very different from 

 Greenland halibut, these two species have some prey 

 in common. In addition, both species are active, off- 

 bottom feeders which could be expected to be more 

 similar in terms of metabolic rates than benthic 

 feeding, small mouth flounders whose food intake 

 has been studied more extensively. The intercept 

 "a" in Equation (2) was 0.0143 for juvenile walleye 

 pollock and 0.0079 for squid. For this study, the 

 intercept for walleye pollock prey was used to cal- 

 culate daily ration when fish was the main prey 

 (>70% of diet by weight), and the intercept for 

 squid was used when squid was the main prey. If 

 the diet was split evenly between fish and squid 

 prey, daily ration was calculated using both inter- 

 cept values to obtain a likely range of daily ration 

 values. 



Average bottom temperatures for the eastern 

 Bering Sea for this study were estimated from 

 oceanographic data on the Bering Sea (Ingraham 



677 



