unaccompanied by other substantial cohorts 

 (Figure 16b); 2) the converse, when November- 

 December 1955 biomass of 10-mm size (together 

 with 11-12 mm, Figure 16c) increased extremely 

 but the percent increase did not keep pace because 

 of strong survival from extended July-September 

 recruitment, seen as piling up in December across 

 8-12 mm range. 



Rate of growth (body length) was seen, above, to 

 be generally steady (Figure 12). Slowed growth 

 was commonest when adolescence or late adult- 

 hood took place during fall-winter. Exceptionally 

 high biomass of 10-12 mm sizes in 1955 and 1956 

 was attributed to greatly slowed growth of 

 adolescents of large cohorts during November- 

 December of both years. 



Regular, less extreme peaking of biomass at the 

 four body lengths just descibed as prominent may 

 be interpreted in terms of differing survival rates 

 among life phases: 



If body-length growth is steady during a given 

 life phase, such as the larval period, biomass 

 growth would proceed as the cube of body length, 

 while population size would be expected to decline 

 exponentially. This inequality leads to a biomass 

 peak at a particular body length which depends on 

 survival rate (Figure 18a). A survival rate of about 

 24%/mo for the larval phase is found to yield such a 

 peak at 4 mm length in the biomass on body-length 

 distribution, a size at which biomass regularly 

 peaks during E. pacijica development. 



Other survival rates were extrapolated from a 

 cluster of age-density curves so as to yield biomass 

 peaks which coincide with real average peaks 

 shown in Figure 16: 43%/mo was found to peak at 7 

 mm, 54%/mo at 10-11 mm, and 66%/mo at 15 mm. 

 A derived age-biomass distribution, linear scale 

 (Figure 18b), is composed of segments based on 

 the above sequence of survival rates. Segments 

 end at 5.8 mm (end of larval phase), 9.3 mm (end of 

 juvenile phase), and 13.2 mm (start of intensive 

 reproduction, after Figure 21b). 



The derived distribution is similar in shape to 

 the observed average annual biomass distribu- 

 tions for 1953 and 1954 (Figure 16a, b). (Growth 

 rates of 1953 cohorts were relatively steady, 

 Figure 12. Those of 1954 appeared less steady but 

 were still without the massive November- 

 December pile-ups of adolescents noted in 1955 

 and 1956.) However, except for the larval period 

 for which the derived and observed mean survival 

 rates (from Figure 14) were both about 23%/mo; 

 other derived rates had to be different from the 



754 



lo- 



co 



CO 

 LiJ 

 CC 

 C3 



g 10 



> 



ir 



z) 

 (/) 



tr 

 O 



t 10 



E 

 O 



o 

 o 



CC 



6 



z 



01 



FISHERY BULLETIN: VOL. 74, NO. 4 



> 



q: 



cj 



cr 

 o 



e 



o 



o 



01 o 



001 



o 



0001 



Body Length 

 3-5Bmm .K 



End 



Larvol Phose 



Survi»ol22%/mo ^ 5B-93mni ^^ 



*3^ . 93-l32mm 



Begin Gonod O«v«lopm«nt 



O 100 



^ Begin Intensive Reproduction 



13.2 mm - 



_L 



_L 



_L 



J_ 



- 2 



E 

 o 

 o 

 o 



in 



CO 



< 



O 



m 



4 6 8 



AGE (MONTHS) 



I I I \ \ L 



10 



II 15 



BODY LENGTH (mm) 



19 



Figure 18. -Hypothetical age-frequency and age-biomass dis- 

 tributions of Euphausia pacijica assuming uniform body-length 

 growth, a, Constant survivorship at each of four rates, selected to 

 yield biomass peaking at 4, 7, 10, and 15 mm, respectively, b, An 

 approximation of annual length-biomass distributions shown in 

 Figure 17, obtained by changing survivorship at life-phase 

 change. 



mean observed rates so as to yield the observed 

 peaks at 7, 10-11, and 15 mm length. These were 

 lower by 24% and 10% for the juvenile and young 

 adult phases respectively, and higher by 6% for the 

 14-18 mm sizes. This means that after the larval 

 phase observed, mean survivorship decreased 

 phase-to-phase by about 4%/mo, whereas in the 

 derived distribution it increased by 11-12%/mo at 

 phase change. This is attributed to deviations 

 from evenness in real growth rates. However, 

 there is a tendency toward progressively positive 

 inflexion with age in certain of the survivorship 

 curves of individual cohorts (Figure 15b-g). 



