FISHERY BULLETIN: VOL. 85, NO. 1 



of these species are much larger than nehu. The 

 reproductive size range of nehu overlaps slightly 

 with only Stolephorus heterolobus and Anchoa naso. 

 Unfortunately, previous studies of these two species 

 involved very few specimens, and the summary 

 statistics must be regarded as less reliable than 

 those of the other species in Table 3. 



Mean relative fecundities for nehu appear to be 

 lower than those of most species; however, the use- 

 fulness of this parameter is questionable because the 

 exponents of the power curves relating fecundity 

 and weight are considerably (and significantly) 

 greater than one in most of the species. Thus mean 

 relative fecundity, a commonly used comparator, 

 would be affected by the size range and size com- 

 position of the sample of females upon which fecun- 

 dity and weight are based. When two groups of 

 similar size composition are compared, as in the case 

 of summer and winter nehu, the difference in mean 

 relative fecundity is similar to that indicated by com- 

 parison of power curves, but otherwise, such as 

 when comparing different-sized species, mean 

 relative fecundities are likely to give erroneous or 

 at best misleading results. Mean relative fecundity 

 also ignores the differences between small and large 

 individuals of the same species or population. 



The exponents of the power curves for nehu are 

 considerably higher than those of any other species. 

 Although the 95% confidence limits for these values 

 do not exclude those for all the other populations, 

 this indicates that the rate of increase in relative 

 reproductive output with increasing size is greatest 

 in nehu. The consequences are illustrated by the 

 relative fecundities calculated for the smallest and 

 largest fish of each population using the power curve 

 for that species (Table 3). Relative fecundities of the 

 largest females are 1.2-2.2 times those of the small- 

 est in the other species but 2.8 and 3.7 times greater 

 in winter and summer nehu, respectively. Both the 

 smallest and largest winter nehu appear to be less 

 fecund per unit weight than the smallest and largest 

 females of all or most of the other species. Small 

 summer nehu also have considerably lower relative 

 fecundity than most of the others, but the value for 

 large summer nehu is among the highest. Ignoring 

 the rather questionable results for Anchoa naso (only 

 12 individuals), the value for the largest Cetengraulis 

 mysticetus is the only one substantially greater than 

 that of the largest summer nehu. 



Although these comparisons must be regarded as 

 tentative because many between-species differences 

 in power curve exponents are not significant, nehu 

 seem to be distinguished from other anchovies not 

 by differences in relative fecundity but rather by dif- 



ferences in the relation between relative fecundity 

 and size. Speculation about the possible relation of 

 this to differences in environment and other life 

 history parameters, such as nehu's short life span 

 and maturity soon after metamorphosis, is un- 

 warranted without evidence that similar differences 

 exist between large and small species in other taxa. 

 Nevertheless, it seems possible that the pattern of 

 allocation of resources between growth and repro- 

 duction over the reproductive life span is yet another 

 life history parameter which could be selected for 

 by prevailing adult mortality rates, predictability of 

 larval survival, etc. 



Comparison of fecundities alone does not ade- 

 quately reflect differences in reproductive effort if 

 there are differences in egg size. For example, nehu 

 eggs average about two-thirds the egg weights 

 calculated for E. mordax by Hunter and Leong 

 (1981). Effort per batch would be best measured by 

 relative cost in terms of dry weight, calories, etc., 

 rather than numbers of eggs. Available data per- 

 mit only crude comparisons of the two species. 



The intercept of the regression equation for G/S 

 vs. fecundity of nehu with ova >0.75 mm is about 

 2.5% for fish from both seasons and nearly the same 

 as the mean G/S (2.4%) of 21 other fish whose 

 largest oocytes were 0.48-0.65 mm and had presum- 

 ably just spawned. (G/S data were not available for 

 fish used for POF analyses.) Using 2.5% as the mean 

 G/S 2 days before spawning and subtracting this 

 from mean G/S of nehu with ova >0.75 mm, i.e., 

 those about to spawn, gives mean relative weights 

 per batch of 3.8% of bodily dry weight in summer 

 and 2.3% in winter. These estimated relative costs 

 per batch are minimal since they do not include in- 

 vestment in bringing oocytes to the size at 2 days 

 before spawning. 



Hunter and Leong (1981) did not give relative cost 

 per spawning of E. mordax in terms of dry weight, 

 but data in their table 4 plus an assumption of dry 

 bodily weight equal to 25% of wet weight yield an 

 estimate of about 4.4% of bodily weight per spawn- 

 ing for an average female. Hunter and Leong' s data 

 in table 1 indicated that dry weight in E. mordax 

 declined about 30% during the main spawning 

 season due to loss of fat; this loss is shown to be 

 equal to the calories required for about 13 spawn- 

 ings. If this is also true for dry weight then the loss 

 per batch would be about 2.3% of dry bodily weight. 



The above estimates of cost per batch in terms of 

 dry weight are very crude and only indicate that 

 nehu, particularly summer nehu, are probably 

 similar to E. mordax. Additionally it is clear that 

 nehu, like E. mordax, lose half or more of their ovary 



136 



