Grover Feeding ecology of Theragrs chalcogramma in the Gulf of Alaska 



467 



other hand, the contribution of Pseudocalanus cope- 

 podids and copepod naupHi decreased with increasing 

 fish size. Other major prey categories (i.e., small cala- 

 noid copepodids, invertebrate eggs, A. longiremis, and 

 Oithona sp.) showed no clear relation to fish size. 



Diet 3S related to site of collection 



Pseudocalanus sp. was the most important prey at 6 

 of the 10 stations, accounting for over 30% (IRI) of the 

 diet at each of these locations (Table 3, Fig. 3). Small 

 calanoid copepodids were of secondary importance, 

 ranging from 13.2 to 23.3%, while invertebrate eggs, 

 copepod nauplii, Pseudocalanus copepodids, and A. 

 longiremis and Oithona sp. each contributed more than 

 10% to the diet at one or more of these stations. 



At two of the remaining four stations Oithona sp. 

 was the primary prey (22.7-25.7%). Small calanoid 

 copepodids were most important at another staion 

 (37.8%i), and A. longiremis was dominant at the re- 

 maining one, comprising more than 53% of the diet. 

 Additionally, invertebrate (mostly copepod) eggs, cope- 

 pod nauplii, pteropods, Pseudocalanus sp., Pseudocala- 

 nus copepodids, A. longiremis, Eucalanus bungii, and 

 small and large unidentified calanoid copepodids (i.e., 

 >1.5 mm), were well represented in the diet at one or 

 more of the stations that were not dominated by Pseu- 

 docalanus sp. 



Diet overlap between stations 



Values of the index of proportional similarity (PS) 

 varied widely, franging from 0.252 to 0.833 (Table 4). 

 Of 11 pairs of adjacent stations, 45% showed no sig- 

 nificant diet overlap, while of 34 pairs of non-adjacent 

 stations 50% showed no significant overlap. Two sta- 

 tions, 30 and 40, were quite dissimilar from all other 

 stations, and station 32 showed significant overlap only 

 in relation to stations 21 and 58. 



Wlien examining diet overlap for the six stations that 

 comprise the transect parallel to the Alaska Peninsula, 

 a high degree of variation was seen in PS values (Table 

 5). At the southwestern end of the transect the lowest 

 PS value was obtained for stations in closest proximity 

 (21 and 30), while at the northeastern end of the tran- 

 sect the greatest PS value was obtained for stations 

 in closest proximity (58 and 44). At the same time, diet 

 at the two most distanct stations (21 and 58) showed 

 great similarity. 



Along one of the two transects perpendicular to the 

 Alaska Peninsula (stations 44, 45, and 48), PS values 

 indicated considerable overlap (Table 6). In contrast, 

 a different pattern was obtained along the other per- 

 pendicular transect (stations 38, 40, and 42) where PS 

 values were wide-ranging. For each of the two tran- 

 sects the greatest diet overlap was between the near- 

 shore and offshore exti'emes of the transect (i.e., sta- 

 tions 44 and 48, and stations 38 and 42). 



