448 



Fishery Bulletin 104(3) 



Overlap between age classes of pollock consumed 

 by northern fur seals and pollock caught 

 by the commercial trawl fishery 



Age-class distributions of walleye pollock in scat and 

 spew samples were compared with pollock size-compo- 

 sition data in commercial midwater trawls (National 

 Marine Fisheries Service, North Pacific Groundfish 

 Observer Program). Trawl data included samples col- 

 lected from August through September 1990 to 2000 

 in fishery management areas surrounding the Pribilof 

 Islands (Berger^; Fig. 1) that encompass northern fur 

 seal foraging habitat (Robson et al., 2004). Walleye pol- 

 lock fork length from trawls was converted to age class 

 by following the method of Sinclair et al. (1994) for 

 determination of overlap with the age class of pollock 

 consumed by northern fur seals (as estimated from scat 

 and spew samples). 



Results 



Prey abundance 



A total of 1127 scats and 204 spews from St. George 

 Island and 2317 scats and 63 spews from St. Paul Island 

 contained prey remains (Table 1). Prey species compo- 

 sition was similar between scats and spews for each 

 island, but the relative importance of primary prey 

 species based on F0% and MNI% varied by sample type 

 and island (Table 2). Primary prey species (FO^ ^5%) 

 found in both scats and spews consisted of gonatid group 

 I, Gb-Bm, Gm-Gm, northern smoothtongue (Leuroglos- 

 sus schrnidti). Pacific herring (Clupea pallasi). Pacific 

 salmon (Oncorhynchus spp.), Pacific sand lance (Ammo- 

 dytes hexapterus). and walleye pollock (Fig. 2). 



Significant interactions between sample type and 

 island were found for Gb-Bm, Gm-Gm, Pacific salmon, 

 and walleye pollock (Table 3); therefore the F0% of 

 these prey was compared within island or sample type. 

 The occurrence of Gb-Bm in scats versus spews from 

 St. George Island was significantly different— spews 

 having a higher F0% than scat. The F0% of Gb-Bm 

 in scats versus spews collected from St. Paul Island 

 was not significantly different. The F0% of walleye 

 pollock was higher for St. George Island scats than 

 for St. George Island spews. Pollock had the highest 

 F0% of all prey in St. Paul Island scat and spews but 

 was not significantly different between sample types. 

 The ¥0% of Pacific salmon was significantly higher in 

 spews than scat samples for both islands. The ¥0% of 

 Gm-Gm was significantly higher in St. George Island 

 spews than in St. Paul Island spews. Of the prey for 

 which no interaction between sample type and island 

 was observed, differences in the ¥0% of gonatid group 



2 Berger, J. 2004. NMFS. database. Alaska Fisheries Sci- 

 ence Center, National Marine Fisheries Service, NOAA, 

 7600 Sand Point Way N.E., Seattle, WA 98115. 



I, northern smoothtongue, and Pacific sand lance were 

 significant between islands and between sample types 

 (Table 3). Pacific herring was only different between 

 islands. Year of collection had a significant effect on the 

 occurrence of all primary prey species except gonatid 

 group I and was included in the model as an additive 

 variable. Although scat and spew samples were collected 

 during the same season each year, annual comparisons 

 between sample types were not possible because of low 

 sample size within years (Table 1). 



Variations in MNIVf were observed between sample 

 types and island for Gb-Bm and walleye pollock (Table 

 2). The MNI<7f of Gb-Bm was higher in spews than in 

 scat samples for both islands. Conversely, the MNI<7r of 

 walleye pollock was higher in scats than in spews for 

 both islands. Scat and spew samples from St. George 

 Island had higher MNI% of Gb-Bm than those collected 

 from St. Paul Island. The MNI^ of walleye pollock in 

 both sample types from St. Paul Island was greater 

 than the MNI% of walleye pollock from St. George Is- 

 land in scats and spews. 



Comparisons of prey age and size 



A total of 1755 walleye pollock otoliths were recovered 

 from 173 St. George Island scat samples containing wall- 

 eye pollock and, of these, 1484 were determined to be in 

 "good" or "fair" condition. Of the 152 walleye pollock oto- 

 liths recovered from 27 St. George Island spew samples, 

 105 were graded as "good" or "fair." From St. Paul Island, 

 17,827 of 20,062 walleye pollock otoliths recovered from 

 902 scat samples and 154 of 202 walleye pollock otoliths 

 recovered from 23 spew samples were determined to be 



