KENDALL ET AL.: GROWTH OF LARVAL WALLEYE POLLOCK 



density upon growth (O'Connell and Raymond 

 1970; Houde 1975), and recent studies on patches 

 of larvae co-occurring with prey suggest that en- 

 hanced growth may be observed at high prey con- 

 centrations in the field (Govoni et al. 1985). 

 Search volumes of 5 mm larvae morphologically 

 similar to walleye pollock are about 10 L per day 

 (Laurence 1982); thus with densities of 1 larva in 

 25 L found in Shelikof Strait in 1981, it is possible 

 that density-dependence is important in larval 

 feeding rate and growth. 



Walleye pollock is widely distributed in the 

 subarctic North Pacific. Larval feeding habit 

 studies have been conducted in Uchiura Bay, 

 Hokkaido, Japan (Kamba 1977) and in the south- 

 eastern Bering Sea (Clarke 1978) where the prin- 

 cipal prey has been found to be copepod nauplii 

 with Pseudocalanus spp. becoming increasingly 

 important as the larvae grow. Larval growth in 

 the same areas has been studied by Hayashi et al. 

 (1968) and Nishimura and Yamada (1984) for 

 Hokkaido and by Walline (1985) and Clarke 

 (1984) for the Bering Sea. Growth rates in field 

 collections and laboratory rearing studies have 

 been shown to be quite variable, from about 0.16 

 to 0.37 mm d"i (Bailey and Stehr 1986). 



We conducted a field study to investigate the 

 ecology of larval walleye pollock in Shelikof 

 Strait in May 1983 by locating and sampling the 

 densest patch of larvae. Here we report on 

 growth, feeding habits, and depth distribution of 

 larval walleye pollock we collected. 



METHODS AND MATERIALS 



Field Collections 



An ichthyoplankton survey of 63 stations on a 

 15 nmi (27.8 km) grid southwest of Kodiak Island, 

 AK, was conducted aboard the NOAA ship Chap- 

 man from 21 to 28 May 1983 (Fig. 2). At each 

 station a MARMAP double oblique bongo tow 

 (Posgay and Marak 1980) was made from the sur- 

 face to 200 m (as water depths permitted) with a 

 60 cm bongo net equipped with 505 jxm mesh nets. 

 Flowmeters were mounted in the net mouths and 

 a bathykymograph was used to determine the 

 maximum tow depth and to evaluate the tow pro- 

 file. A neuston net (Sameoto and Jaroszynski 

 1969) with 505 ixm mesh was also towed for 10 

 minutes at each station. The neuston net sample 

 and one of the bongo net samples at each station 

 were preserved in 5% sodium borate buffered for- 

 malin in seawater. Most walleye pollock larvae 



from the other bongo net were rough sorted at sea 

 and were immediately preserved in buffered 90% 

 ethanol for otolith examination. 



The results of the sorting of larvae at sea were 

 used to choose a location of high larval density for 

 the diel feeding/distribution study (Fig. 2). An- 

 other oblique tow, after this survey, confirmed 

 the presence of high concentrations of larvae. 

 Several preliminary tows with four 20 cm bongo 

 nets on the towing wire fixed at 10 m depth inter- 

 vals (between 5 and 91 m) were made to find 

 depths of maximum larval concentrations. A tow 

 was then taken every 4 hours for 48 hours during 

 28-30 May 1983 with 20 cm bongo nets equipped 

 with 253 |xm mesh nets on one side and 333 ixm 

 mesh nets on the other. Four nets were fished 

 simultaneously for 10 minutes at a ship speed of 

 approximately 100 cm/second. The nets were 

 placed on the wire to fish at four depths within 

 the region of larval abundance (nominally 20, 30, 

 40, and 50 m). Flowmeters were mounted in the 

 mouths of the nets, and a bathykymograph was 

 deployed with the deepest net to record actual tow 

 depths. During setting and retrieving, the ship 

 maintained reduced speed to minimize fishing 

 outside the chosen depth strata. Thus, although 

 no closing devices were used, nearly all of the 

 water passing into the nets was at the chosen 

 depth (Kendall and Naplin 1981). Tows were 

 made at 1030, 1430, 1830, 2330, 0230, and 0630 

 local time (sunrise was at 0455 and sunset 2138 

 h). During the sampling of the stations at 1430 

 and 0230 on both days, a 1 m^ mechanical Tucker 

 trawl with 505 |xm mesh was fished for 10 min- 

 utes at 35 m to investigate escapement from the 

 20 cm bongo nets. Also during the second 24-h 

 period a 60 cm bongo net with 505 jxm mesh was 

 fished about 2 m below the deepest 20 cm bongo 

 net to stabilize the wire and allow further catch 

 comparisons. 



Expendable bathythermograph (XBT) casts 

 were done at each survey grid station and at the 

 1430 and 0230 vertical distribution study sta- 

 tions. Conductivity-temperature-depth (CTD) 

 casts (Ocean Data Equipment Corporation'^ 

 Model 302) were made at 15 of the survey grid 

 stations selected to provide three sections across 

 the major southwesterly setting flow field in the 

 area (Fig. 2). 



"Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



501 



