FISHERY BULLETIN: VOL. 80, NO. 3 



chemical composition (Anger and Nair 1979); 

 comparison of both methods with direct calo- 

 rimetry in identical material should be worth- 

 while. However, the J values given in this paper 

 compare favorably with those reported else- 

 where for decapod larvae (e.g., Cummins and 

 Wuycheck 1971; Mootz and Epifanio 1974; 

 Logan and Epifanio 1978; Capuzzo and Lan- 

 caster 1979; Dawirs 1980; Stephenson and 

 Knight 1980). 



The interpretation of changes in the relative 

 chemical composition of larvae (C:N, C:H) leads 

 to interesting assumptions about physiological 

 processes, but there is a need for complementary 

 biochemical investigations. Such analyses are 

 planned for future studies as an extension of the 

 present results. So far, our figures compare 

 favorably with those given in the literature 

 (Childress and Nygaard 1974; Ikeda 1974; Omori 

 1979; Dawirs 1980). Ikeda (1974) investigated a 

 large number of zooplankton species, and he 

 found that C:N (by weight) is in most cases 3 to 5, 

 whereas C:H is mostly 6 to 7. 



Growth (any measure except size and FW) 

 during larval development in decapod Crustacea 

 usually follows — at least for some period — an 

 exponential pattern, if gain from stage to stage is 

 considered (Reeve 1969; Mootz and Epifanio 

 1974; Logan and Epifanio 1978; Johns and 

 Pechenik 1980; Stephenson and Knight 1980). 

 This also holds true for the zoeal stages of Hyas 

 araneus. Interpolation of biomass values within 

 single stages from such exponential curves 

 yields poor correspondence of predicted and ob- 

 served data, because growth within stages 

 follows different patterns. In both zoeae it could 

 be described most accurately by power func- 

 tions, whereas exponential regressions do not fit 

 as well. At the end of the molt cycle, however, 

 such parabola-shaped fitted growth curves lose 

 their applicability. This final period probably 

 corresponds to the stages D 2 to D 4 , during which 

 molt is prepared by separating the epidermis 

 from the old cuticle (Freeman and Costlow 1980). 

 Possibly, there is no more significant food uptake 

 during this phase of body reconstruction (Anger 

 and Dawirs 1981). 



In the megalopa, another growth pattern was 

 found. The period of body reconstruction and of 

 presumed inability to take up food preceding 

 metamorphosis appears to be much longer in this 

 stage. The daily energy loss per individual was 

 three times higher during this time as opposed to 

 megalopae starved from the beginning. This 



contrast suggests that a final fasting period is a 

 normal part of the development program, not 

 considered starvation, and thus not counter- 

 balanced by reduced metabolism. This assump- 

 tion is supported by a number of observations in 

 other decapod megalopae (Mootz and Epifanio 

 1974; Schatzlein and Costlow 1978; Dawirs 1980; 

 for recent review see Anger and Dawirs 1981). 



The duration of the megalopa stage is much 

 more variable than the zoeal instars. This fact 

 may be related to the ecological role of the 

 megalopa which is to select a biotope suitable for 

 adult life. The capacity to delay selection should 

 be related to the amount of reserve accumulated 

 prior to the change in energy balance. This 

 strategy is in contrast to that observed by 

 Pechenik (1980) in gastropod larvae. These do 

 not cease to grow with the onset of metamorphic 

 competence, and their capability to delay meta- 

 morphosis appears to be related to the preceding 

 growth rate. More detailed investigations on the 

 nutritional and ecological needs of the megalopa 

 stage are necessary for a better understanding of 

 this critical phase in benthic recruitment. 



All these complicated changes of biomass as 

 well as their extent (two- to threefold increases 

 within single stages) suggest that the use of 

 "characteristic" values for particular instars ap- 

 plied in energy budgets and other energetic 

 considerations must lead to very rough figures. 



Another complicating factor is annual or 

 seasonal variation in initial biomass of hatching 

 larvae and in their growth rate. Since viability 

 also appears to be related to this kind of 

 variation, future studies will have to examine its 

 degree and significance. The same holds true for 

 possible systematic differences between labora- 

 tory-reared larvae and those obtained from 

 wild plankton. Several authors (Knight 1970 and 

 earlier papers; Rice and Provenzano 1970; Ingle 

 and Rice 1971) observed higher growth rates in 

 naturally grown developmental stages of 

 different decapod species. 



Only a small part of the organic matter 

 accumulated during the zoeal stages is lost in 

 exuviae. This is much different in the megalopa. 

 More than three times more matter and energy 

 was found in its cast exoskeleton than in both 

 zoeae combined. These and other striking dis- 

 similarities between zoea and megalopa larvae 

 underline their different roles. The former 

 accumulate energy-rich substances taken from 

 the pelagic food web, and they are responsible for 

 dispersal of the species. The latter stage, which 



430 



