prey of other cetaceans and numerous birds and 

 fishes. 



Arctic cod are circumpolar in Arctic waters ex- 

 tending south to at least lat. 60° N on the Alaska 

 coast, typically in association with sea ice (An- 

 driyashev 1954). They are a species of key trophic 

 importance upon which many other far northern 

 marine consumers depend entirely for a major 

 portion of their yearly nutritional requirements. 

 They are eaten by at least 12 species of marine 

 mammals, 20 species of birds, and 5 species of 

 fishes. Arctic cod are especially important because 

 in the areas and at the times when they are abun- 

 dant they are the only forage fishes present. 



Investigations of food habits of marine animals 

 almost invariably involve analysis of stomach con- 

 tents. Morrow (1979) published preliminary keys 

 to otoliths of 16 families of fishes found in Alaskan 

 waters including the Gadidae, whereby fishes 

 eaten by predators can be identified from otoliths 

 even after soft parts and bones have been digested. 

 In most instances the size of the fish or meal can 

 also be determined from otoliths through back cal- 

 culation offish length and/or weight from various 

 measurements of otolith size (Morrow 1951; Tem- 

 pleman and Squires 1956; Southward 1962; 

 Gjosaeter 1973). 



In this paper we present relationships of otolith 

 length to fish length and weight for pollock, saf- 

 fron cod, and Arctic cod of the Bering, Chukchi, 

 and Beaufort Seas. 



Methods 



Samples of fishes were obtained by otter trawl- 

 ing in the Bering, Chukchi, and Beaufort Seas 

 (Table 2). Soon after capture all fishes were iden- 

 tified, weighed to the nearest 0.1 g, and fork length 

 measured to the nearest millimeter. The sagittal 

 otoliths were removed and length and width mea- 

 sured to the nearest 0.1 mm with vernier calipers. 



When otolith lengths and widths were plotted 

 against fish lengths as scatter diagrams, the rela- 

 tionship between otolith length and fish length 

 was found to be less variable than that of otolith 

 width and fish length. For this reason otolith 

 length was taken as the criterion for otolith size 

 and used in subsequent calculations. Casteel 

 (1976) discussed in detail the reasons for using 

 length as the best measure of otolith size. 



We chose a double regression method for relat- 

 ing otolith size to fish size (Fitch and Brownell 

 1968; Casteel 1976). For each species the relation- 

 ships of otolith length to fish length and fish length 

 to fish weight were calculated. In cases where two 

 equations were required to fit a single relation- 

 ship, the inflection point was determined by itera- 

 tion. The specified inflection point was varied 

 by increments of 0.1 and the pair of equations 

 which minimized the combined deviation was 

 selected. 



Results and Discussion 



Regressions offish fork length on otolith length 

 differed markedly among the three species. Those 

 of walleye pollock and saffron cod formed two dis- 

 tinct straight-line sections each, with inflection 

 points at otolith lengths of 10 mm in walleye pol- 

 lock (fish length 22 cm) and 8.5 mm in saffron cod 

 (fish length 15 cm) (Figures 1, 2). The regression 

 for Arctic cod was rectilinear over the range of 

 samples (Figure 3). 



Several sources of error are possible when es- 

 timating the size of a fish from its otoliths, among 

 which are normal variability in the ratio of fish 

 length to otolith length and differences in lengths 

 of left and right otoliths of the same fish. The 

 calculated regression coefficients show that such 

 variability is quite small. Deviation between ac- 

 tual measured and calculated fish lengths was 

 usually <5%. Since food habits studies deal with 



Table 2. — Sources of Alaskan marine gadids measured to determine otolith length-fish size relationships. T = Theragra 



chalcogramma; E = Eleginus gracilis; B = Boreogadus saida. 



Vessel and cruise no. 



Date 



Area 



Depth range (m) Trawls (no.) Species 



NOAA' Ship Surveyor (RP-4-SU-76AI&II) 



NOAA Ship Discoverer (RP-4-DI-76BIII) 



USCGC2 Glacier (AWS76) 



NOAA Ship Miller Freeman (RD-4-MF-76BII 



NOAA Ship Surveyor (RD-4-SU-77AII, III) 



NOAA Ship Discoverer {RD-4-DI-77AVI) 



NOAA Ship Surveyor (RD-4-SU-77BII) 



USCGC Glacier (AWS77III) 



ADF&G3 skiff (Shishmaref 78) 



NOAA Ship Surveyor (RP-4-SU-78AV VI) 



'National Oceanic and Atmospheric Administration. 

 ^United States Coast Guard Cutter, 



^Alaska Department of Fish and Game. 



189 



