650 



Fishery Bulletin 103(4) 



years were pooled for each month. Abundance did not 

 appear to affect the spatial distribution of eggs or lar- 

 vae; their distribution patterns were similar no matter 

 whether abundance was high or low. Months of highest 

 abundance (April, May, and June) were divided into 

 early to mid-month (days 1-15) and mid- to late month 

 (days 16-31). The area covered by the cruises was di- 

 vided into 50x50 km grid cells. Mean catch per cell was 

 calculated for each grid cell, averaging over all stations 

 falling within the cell. For the areas other than Shelikof 

 Strait, the number of stations per cell ranged from 1 

 to 10. The most intensive sampling was conducted in 

 the Shelikof Strait area, south to approximately 56°N 

 latitude. Cells in this area, depending on the month, 

 could have more than 100 stations within them. To ex- 

 amine larval drift, the center and ellipse (centroid) of 

 egg and larval abundance for early and late May 1994 

 and 1996 (two years with different flow regimes in She- 

 likof Strait) were calculated according to the methods 

 described in Kendall and Picquelle (1989). 



The vertical distribution of eggs and larvae was exam- 

 ined from samples from four 1-m 2 MOCNESS (multiple- 



opening-closing-net and environmental sensing system) 

 tows. For each tow, 6 to 8 depth intervals were sampled 

 from near the sea floor to near the surface. The samples 

 were collected during peak spawning in 1991 (one tow 

 during day light ), 1993 (one tow during day light), and 

 1996 (two tows: 1996A conducted during the night, and 

 1996B during day light). Eggs collected in each depth 

 interval were categorized as early (stages 1-12), middle 

 (stages 13-15), and late stage (stages 16-21) according 

 to walleye pollock egg stages adapted from Blood et al. 

 (1994). The late stage was divided into two categories: 

 late A (stages 16-19) and late stage B (stages 20-21), 

 to indicate which eggs were closest to time of hatching. 

 Taking into account shrinkage in standard length due 

 to collection and preservation (Theilacker and Porter, 

 1995), larvae were divided into three size categories 

 based on development. Larvae <5 mm were classified 

 as recently hatched, larvae 5-6 mm as prefeeding or 

 first feeding, and larvae >6 mm as feeding, based not 

 only on size but also on the amount of yolk present, and 

 whether prey were visible in their gut. These catego- 

 ries were based on observations of fiathead sole larvae 



