Kendall et al .: Vertical distribution of eggs and larvae of Theragra chalcogramma 



543 



Alaska Peninsula 



=t 



Cape 



Kekurnoi >-* , ' + 1-1 



^*) / +-2-1.5.6,7.3,8 

 1-3 + 

 / +2-2.3.4 



+ + 5 



t, + +6 



4 , 



Sutwik I. 



+5 ' 



+7 ' 



Semidi Is. . 



f/ Trinity Is 



6 



Figure 1 



Locations of MOCNESS sampling series ( large numbers 

 tows (small numbers). Tows are shown only when there 

 substantial differences among their locations. 



centrations, partially because sampling then was 

 southwest of the main spawning area (Fig. 1), and 

 partially because it was after peak spawning. The 

 notable exception was the single tow of series eight 

 taken in the area of maximum spawning. Although 

 it was mid-May, egg densities were relatively high 

 ( 139/1,000 m 3 ), indicating that some spawning had 

 occurred within the previous two weeks. 



Larvae were abundant in all series except the first 

 when they were absent (Table 2). Mean density of 

 larvae among the tows in series two through nine 

 ranged from 39 to 509/1,000 m 3 (SD ranged from 54 

 to 1,011/1,000 m 3 ). Series were not always in the ex- 

 pected area of maximum concentration of larvae and 

 thus do not necessarily represent the seasonal trends 

 in larval abundance (see Kendall and Picquelle, 

 1990). 



Series four followed a surface drifter with a drogue 

 at 30-35 m (Incze et al., 1990), whereas sampling 

 during series nine was at a fixed geographic loca- 

 tion. During the 2.5 days of series four, the buoy 



moved anticyclonically. Catches in these two 

 series varied considerably; during series four 

 the mean density among tows was 82-285 lar- 

 vae/1,000 m 3 , compared with 42-482 larvae/ 

 1,000 m 3 during series nine (Table 2). The coef- 

 ficient of variation of density among tows for 

 series four was 0.63 and for series nine it was 

 0.68, indicating that variability among tows 

 using the two sampling strategies was similar. 



Overall depth distributions 



Mean depths of eggs decreased during the sea- 

 son. Multiple comparison tests of mean depths 

 of eggs showed significant differences between 

 series one, two, and three, when eggs were most 

 abundant. Among the 10 tows in the first two 

 series, the observed mean depth of eggs was 

 between 153 and 206 m (Fig. 2). In series three 

 through eight, the observed mean depth was 

 less than 130 m, but the number of eggs was 

 relatively small. The shallower towing schemes 

 of series four through eight may have biased 

 the mean depth of eggs, but the general trend 

 is thought to be real. 



During the second series, when only newly 

 hatched larvae were present, their observed 

 mean depths of occurrence were from 165 to 

 212 m among tows (Fig. 3). One standard de- 

 viation of mean depth was 27 to 73 m and gen- 

 were erally increased during the series (Fig. 3). In 

 the third series, when recently hatched larvae 

 dominated, larval mean depths varied from 70 

 to 106 m (range of SD: 83-91 m). As opposed to 

 series two and three, when larvae were mainly found 

 below 100 m, mean depths of larvae during series 

 four ranged from 24 to 58 m (range of SD=15-71 m) 3 . 

 Mean depths of occurrence of larvae from series five 

 through seven (13-18 May 1986) varied from 15 to 

 47 m (range of SD=8-36 m) (Table 3). In series eight, 

 taken in mid-May in the spawning area, the larvae 

 averaged 4.6 mm (SD: 0.18 mm) (Table 2), and their 

 mean depth of occurrence was 21 m(SD= 18 m). Dur- 

 ing series nine in late May, mean depths of larvae 

 among the tows ranged from 15 to 38 m (Table 3) and 

 varied on a diel basis (see below). 



Larvae in the noon tow on the second day of sampling had a 

 mean depth of 58 m (SD=71 m). This was due to an unusually 

 large catch in the deepest net (607/1,000 m 3 [23% of all the lar- 

 vae in the towj at 150-200 m) of larvae with a mean length of 

 4.71 mm. This appeared to be larger than the overall mean of 

 the larvae collected at this depth during this series (4.35 mm), 

 indicating the catch was not all newly hatched larvae that had 

 not moved to the upper water column. If we discount this net, 

 the mean depth of larvae in this tow was 21.5 m, close to the 

 value in the other tows of the series (Fig. 4). 



