Grover Feeding ecology of Theragra chalcogramma in the Gulf of Alaska 



469 



Table 6 



Index of proportional similarity (PS) for walleye 

 pollock. 15.0-19.9 mm SL. along two transects 

 perpendicular to the Alaska Peninsula, calcu- 

 lated using mean volume data. PS values for 

 adjacent stations are underlined. 



Discussion 



The portion of the present study dealing with the rela- 

 tionship between diet and fish size complements the 

 study by Kendall et al. (1987) conducted in the Gulf of 

 Alaska during May 1983. As the samples in 1983 were 

 collected earlier in the year using different gear (a 

 505-^m mesh bongo net) than in 1987, the size range 

 of larvae was substantially different in the two studies. 

 However, a comparison of the diet of larvae 10.0-19.0 

 mm can be made between the two years. From this it 

 was clear that copepod eggs and nauplii were of some 

 consequence in the diet in both years. A between-year 

 comparison of the relative importance of Pseudocala- 

 nus sp. adults and copepodids cannot be made because 

 the 1983 dietary analysis considered adult and imma- 

 ture stages together as copepodids (CI-CVI). In both 

 years Pseudocalanus sp. was the dominant copepod, 

 followed by Acartia sp. and Oithona sp. As Pseudo- 

 calanus sp. carries its eggs (Corkett and McLaren 

 1978), the majority of copepod eggs in the diet were 

 probably incidentally ingested attached to adult 

 females. However, berried females were not observed 

 in pollock stomachs. While the delicate structure of 

 Pseudocalanus egg sacs may preclude larval fish from 

 ingesting females with eggs intact (A.J. Paul, Univ 

 Alaska, Seward, pers. commun., Nov. 1988), it is also 

 possible that some of the eggs that were ingested could 

 have been from other genera such as Acartia or Cala- 

 nus that broadcast their eggs into the sea (C.B. Miller, 

 Oregon State Univ., Corvallis, pers. commun., June 

 1989). The source of copepod eggs cannot be resolved 

 without observing female copepods with eggs in the 

 diet or identifying eggs to species. Neither condition 

 was met in the present study. 



The dietary importance of copepod nauplii decreased 



and Pseudocalanus minuius increased with increasing 

 larval size for larvae 9-20 mm SL in Uchiura Bay, Hok- 

 kaido (Kamba 1977), as was the case in the present 

 study. However, while /I. longiremis, Oithona sp., and 

 various copepodids were important in the Gulf of Alas- 

 ka, they were not in Uchiura Bay. Since copepod eggs 

 were excluded from the dietary analyses by Kamba, 

 it is not possible to compare their relative importance 

 in Uchiura Bay with the Gulf of Alaska. In the Bering 

 Sea, Clarke (1978) reported that copepod nauplii and 

 Oithona similis were the primary prey for pollock 9-18 

 mm SL, and Cooney et al. (1980) reported that cope- 

 podids (CI-CVI) and copepod nauplii were dominant. 

 Oithona similis was the dominant copepod species, 

 followed by Pseudocalanus sp. and A. longiremis. In 

 both Bering Sea accounts, copepod eggs accounted for 

 approximately 5% of the diet of this size class, less than 

 in the Gulf of Alaska. 



While the diet of 20.0-29.9 mm SL fish in the Gulf 

 of Alaska appeared to be more diverse than in Uchiura 

 Bay (Kamba 1977), in both areas Pseudocalanus sp. was 

 the dominant prey, and Calanus sp. was well repre- 

 sented in the diet. In the Bering Sea Pseudocalanus 

 sp. copepodids were the dominant prey and copepod 

 eggs were also very important in the diet of early 

 juveniles, although a broader size class, 20.1-60.0 mm 

 SL, was utilized (Cooney et al. 1980). In a more recent 

 study, also in the Bering Sea, cyclopoid copepods 

 proved to be the primary prey for fish 20.0-29.9 mm 

 TL, with Pseudocalanus sp. adults second in impor- 

 tance (Lee 1985). In the Gulf of Alaska, Oithona 

 sp. (the dominant cyclopoid) was of highly localized 

 importance. 



The manner in which diet of larval and juvenile 

 pollock varied between different locales and between 

 years in the same locale illustrates how diet may vary 

 in response to available food resources, which vary in 

 response to oceanographic conditions (e.g., Grover and 

 011a 1987). As plankton data for the present study were 

 unavailable, the extent to which diet resulted from prey 

 selectivity cannot be determined. However, assuming 

 that diet was at least reflective of available food re- 

 sources, the wide range in PS values would indicate 

 that food resources were not homogeneous. Diet varied 

 widely from station to station, with nearly as many non- 

 adjacent as adjacent station pairs showing significant 

 dietary overlap. This type of variation would suggest 

 areas of planktonic patchiness, perhaps reflecting 

 underlying hydrographic heterogeneities. Low PS 

 values may indicate small prey patches, while high 

 values that were seen across several stations may be 

 indicative of large prey patches. 



Regardless of whether prey were energetically equi- 

 valent at each station, the incidence of feeding was 



