Gabr et al.: Reproductive versus somatic growth during the life cycle of Sepia pharaonis 



805 



(i.e. mantle, head digestive gland and viscera masses) 

 with mantle length (P<0.05). The effect of ML on re- 

 productive organ mass (i.e. ovary, nidamental gland 

 masses) also was significant (P<0.05). In relation to 

 nidamental gland length, there was a small but sig- 

 nificant decrease (P<0.05) in total somatic mass, 

 mantle mass, and head mass, whereas the relative 

 decrease in digestive gland mass and viscera mass 

 was not significant (P>0.05). As expected, there was 

 also a highly significant increase (P<0.05) in repro- 

 ductive organ mass with maturity (NGL). 



The second analysis was an analysis of multivari- 

 ate covariance (ANCOVA). In an animal that matures 

 at a wide range of body sizes, it is difficult to quan- 

 tify the pattern of maturation because comparisons 

 of individuals of different sizes cannot be thought of 

 as indicating the pattern of growth and maturation 

 of any one animal. Thus, although some animals 

 matured at 60 mm ML, others remained immature 

 at larger sizes ( Fig. 1). ANCOVA corrects for the varia- 

 tion in size (ML) of animals within maturity stages. 



Table 2 shows the initial standard ANCOVA de- 

 sign to test for homogeneity of slopes. In females, 

 there was significant effect of an interaction between 

 the factors (maturity stage) and covariate (mantle 

 length) for OM and NGM. The effect of interaction of 

 mantle length and maturity stages on TSM, MM, 

 HM, DGM, and VM was not significant (P>0.001). 

 Thus in these cases the hypothesis of homogeneity 

 of slopes was accepted; therefore further analyses 

 were required to test the effect of maturity stages 

 for these variables. In males, there was a highly sig- 

 nificant effect (P<0. 001) between mantle length and 

 maturity stage for all log-transformed dependent 

 variables except spermatophoric complex mass. Thus 

 for these variables the hypothesis of homogeneity of 



slopes was rejected and further analysis was not pos- 

 sible (i.e. the standard ANCOVA design should not 

 be developed for these cases). 



The results of the final ANCOVA design (in the 

 cases where the effect of mantle length and matu- 

 rity stage interactions were not significant) are indi- 

 cated in Table 3. In female Sepia pharaonis, matu- 

 rity significantly affected TSM, MM, HM, and VM 

 (P<0.001) but did not significantly affect DGM 

 (P>0.001 ). In males, maturity significantly influenced 

 spermatophoric complex mass (P<0.001). 



Figure 2 shows the adjusted means for each de- 

 pendent variable for each maturity stage after the 

 removal of the mantle length (covariate) effect when 

 the effects of maturity stages were found to be sig- 

 nificant. For females, there was a decrease in total 

 somatic mass, mantle mass, and head mass with in- 

 creasing maturity stage. There was a decrease in 

 viscera mass from stage I to stage III followed by an 

 increase at stage IV (Fig. 2A). For males, there was 

 an increase in relative mass of spermatophoric com- 

 plex mass with increasing maturity stage (Fig. 2B). 



Monthly variation in soma versus gonad production 



Monthly fluctuations in mean dry weight indices for 

 individual organs of somatic tissue (mantle index, 

 MI; head index, HI; digestive index, DGI; and vis- 

 cera index, VI) and reproductive tissue (gonadoso- 

 matic index, GSI; nidamental gland index, NGI; and 

 spermatophoric complex index, SCI) for females and 

 males are illustrated in Figures 3 and 4, respectively. 

 These indices displayed monthly fluctuations and 

 indicated the relative proportion that each body com- 

 ponent contributed to the weight of the entire body. 

 The mantle mass was always the largest component 



