552 



Fishery Bulletin 92(3). 1994 



lert et al., 1985; Yamashita et al., 1985; Sogard 

 et al., 1987; Davis et al., 1990). Fewer studies have 

 examined the vertical distribution offish larvae dur- 

 ing crepuscular periods. The present study has shown 

 that larger larval pollock range deeper during the 

 day than at night and that, at dawn and dusk, they 

 are present at shallower depths in the water column 

 than at midday. We hypothesize that these changes 

 in vertical position allow pollock larvae to extend the 

 length of their daily feeding period. 



In the laboratory, first-feeding pollock larvae could 

 not feed at light levels below 0.006 |imol-m _2 -s -1 (Paul, 

 1983). Except at night, light levels are brighter than 

 those at the depths where we found feeding-stage 

 larvae. In studies of behavioral responses of walleye 

 pollock larvae (4-8 mm SL) to light in the labora- 

 tory, Olla and Davis (1990) found reduced activity 

 and orientation in a nonfeeding mode at light levels 

 <0.01 and avoidance of light at levels 

 >13 u.mol-m~ 2 s _1 . In the dark, larvae migrated up- 

 ward and remained in the upper part of the cham- 



I 3 5 7 9 11 13 



Sampling Day 



Figure 1 1 



Predicted light levels at depth on days of sampling and mean 

 depths of walleye pollock, Theragra chalcogramma , larvae in 

 MOCNESS tows ( indicated by circles for daytime tows, triangles 

 pointed down for dusk tows, asterisks for night tows, and tri- 

 angles pointed up for dawn tows). Depths of three light levels 

 are plotted: 50, 10 and 0.01 umol-m~ 2 s _1 . Light levels above 50 

 are clear, those 10-50 are light gray, those 0.01-10 are medium 

 gray, and those less than 0.01 |imol-m ' 2 s _1 are dark gray. Light 

 levels are based on incident light at the Kodiak airport during 

 the sampling periods with an extinction coefficient of 0.16. 



bers, demonstrating negative geotaxis or barotaxis, 

 or both. Light levels between 0.01 and 10 umolirr 2 -s _1 

 are estimated to have occurred between about 25 and 

 60 m during our sampling in Shelikof Strait. The 

 mean depths of feeding larvae were typically in the 

 upper part of this range (Fig. 11). Larvae longer than 

 7 mm seemed to adjust their vertical position on a 

 diel cycle to stay at light levels similar to those "pre- 

 ferred" in the laboratory. At night these larvae were 

 present at depths where light had been greater than 

 10 p:mol-nr 2 -s _1 during the day. 



The relationship of vertical distribution of larval 

 fish to vertical temperature structure of the water 

 column varies among species (Kendall and Naplin, 

 1981; Sogard et al., 1987). Hypothetically, there are 

 metabolic advantages to diel descents into cooler 

 waters ( Lampert, 1989). Larvae that stay nearer the 

 surface (at higher temperatures) at night when they 

 are digesting their food may accrue such advantages 

 (Wurtsburgh and Neverman, 1988). However, given 

 the small differences in temperature with depth ob- 

 served here (~1°C), energetic advantages are almost 

 certainly insignificant compared with the ad- 

 vantages of feeding at optimal light levels and 

 at depths of maximum prey abundance. An 

 alternative advantage of residing deeper, and 

 thus at lower light levels during daytime, 

 may be to avoid visual predators. 



Acknowledgments 



Many people, mainly from the AFSC, helped 

 collect and process the samples and data upon 

 which this paper is based. We thank them 

 all, particularly those who endured the harsh 

 conditions of Shelikof Strait during the 

 cruises there. David Nieman rendered his 

 expert help in operating the MOCNESS on 

 several cruises. Allen Macklin (PMEL) is 

 thanked for obtaining and analyzing meteo- 

 rological data presented here. Richard Davis 

 (AFSC) graciously let us use irradiance with 

 depth data obtained by his modelling efforts. 

 An earlier draft of the manuscript was helped 

 immeasurably by reviews of many of the re- 

 searchers involved in the FOCI (Fisheries- 

 Oceanography Coordinated Investigations) 

 program: Kevin Bailey, Jim Schumacher, Jeff 

 Napp, Ric Brodeur, Bori Olla, Gary Stauffer, 

 and Susan Picquelle (whom also aided in sta- 

 tistical analyses). Ian Perry (Canada, Depart- 

 ment of Fisheries and Oceans) also provided 

 a very helpful review of the manuscript. 



