FISHERY BULLETIN: VOL. 81, NO. 3 



fish, whereas no consistent pattern emerged between 

 large and small Atlantic cod. 



A significant proportion of the stomachs in the sam- 

 ples was empty, especially among silver hake (Tables 

 3,4). Moreover, among those fish containing measur- 

 able amounts of food in their stomachs, the distribu- 

 tion of stomach content weight was strongly skewed 

 towards small values. For these two reasons, the me- 

 dian stomach content weight was always con- 

 siderably less than the mean weight (Table 5). 



The mean and median stomach content weights 

 within the entire population, which includes fish that 

 had empty stomachs, were less than the correspond- 

 ing values when only those fish containing measur- 

 able quantities of food were considered (last three 

 columns of Tables 3 and 4; last two columns of Table 

 5). The differences were greater for silver hake than 

 Atlantic cod, because of the large percentage of hake 

 with empty stomachs. These data show that, even 

 when the analysis is restricted to the fish that have 

 recently fed (i.e., the fish with nonempty stomachs), 

 the amount of food in the stomachs is, on the average, 

 very much less than the maximum amount that the 

 fish are physically capable of ingesting (as indicated 

 by the maximum observed values, Table 5). 



The mean stomach contents by season for each 

 species, all sizes combined, are presented in grams 

 and as a percentage of body weight in Tables 3 and 4. 

 When the stomach contents are expressed in terms of 

 weight (grams), the apparent importance offish prey 

 in the diet is greater than when the stomach contents 

 are expressed as a percentage of body weight. This 

 disparity occurs because the large, more piscivorous, 

 fish contain a much greater weight of food in their 

 stomachs, and have a disproportionate effect com- 

 pared with the more numerous, but less piscivorous, 

 small fish. This bias is eliminated if the stomach con- 

 tent weight is normalized to the weight of the fish 

 (percent body weight) for the calculation of mean 

 stomach contents. This also allows intercompari- 

 son of samples with different size-distributions of 

 fish. We have, therefore, expressed the stomach 

 content data as percent body weight for the analy- 

 sis of diel changes in stomach fullness and daily 

 ration. 



Diel Changes in Stomach Contents 



The diel feeding patterns of large and small silver 

 hake appeared to differ. The weight of the stomach 

 contents of small hake fluctuated over the 24-h day, 

 but did not show any trends indicative of diel 

 periodicity in feeding intensity during either spring 

 or fall (Fig. 3). However, large hake exhibited a strong 



SILVER HAKE, < 20 cm 

 SPRING FALL 



2.0 



- 1.5- 



5 



0- 



O 



CO 



iLl 



o 

 (_> 



I 

 o 

 < 



o 



H 

 CO 



0.5- 



T, 



zz 



T. 



n 



P 



5.3 



zm 



SILVER HAKE , > 20 cm 



SPRING 



FALL 



2.0- 



15- 



1.0- 



0.5- 



7. 



T, 



m 



J 



m 



24 06 12 18 24 



24 06 12 18 24 



TIME , hours 



FIGURE 3.— Diel changes in "total stomach content weight," "fish 

 prey," and "other prey" in the stomachs of small and large silver 

 hake during spring and fall. 



pattern of nocturnal feeding during the fall (Fig. 3). 

 This pattern was less clear during spring, when com- 

 paratively large amounts of food were observed in the 

 stomachs during the daytime. However, maximum 

 amounts of food occurred at night, between 2100 

 and 0300. 



Among small silver hake, the number of empty 

 stomachs averaged 23.4% during spring over the 24- 

 h day, with highest values observed between 1500 

 and 2400 (see Fig. 5). During the fall, the mean num- 

 ber of empty stomachs was 21.9% , with no apparent 

 diel pattern. Among large hake, the number of empty 

 stomachs averaged 40.4% during spring and 47.8% 

 during fall. The percentage remained fairly constant 

 throughout the day except for peaks between 1500 

 and 2100 during spring and 0900 and 1500 during 

 fall. These peaks corresponded to the periods when 

 the minimum mean weight of stomach content was 

 observed (Fig. 3). 



In contrast to silver hake, there were no apparent 

 diel trends in the stomach contents of Atlantic cod 

 (Fig. 4). The percentage of empty stomachs was 

 lower in cod than in hake, and there was no diel pat- 

 tern (Fig. 5). Among large cod, the percentage of empty 

 stomachs over the day averaged 6.4% during spring 



444 



