FISHERY BULLETIN: VOL. 87. NO. 3, 1989 



Encounter 



Incidental filtering of eggs by encounter of 

 other schools' patches of eggs would appear to be 

 a direct mechanism for parental control of re- 

 cruitment rate and population density. If one 

 combines the information on egg cannibalism 

 rate by Alheit (1987) and the demogi'aphic influ- 

 ence on population annual fecundity reported by 

 Pauly and Soriano (1987) (Table 5), one can see 

 that the effective fecundity (eggs hatched after 

 cannibalism) would be low when the stock was 

 composed primarily of juveniles and first-year 

 spawners. 



Table 5. — Age-specific rates of cannibalism and egg 

 production for anchovy. 



It is also possible for cannibalism to interact 

 indirectly with population density. An indirect 

 method for population control is evident if one 

 considers a system where the anchovy prefer to 

 feed by biting relatively large zooplankton and 

 where filtering occurs only when large zooplank- 

 ton are low in abundance (O'Connell 1972). 

 Under these circumstances predation by an- 

 chovy on the major herbivores could result in a 

 larger standing crop of phytoplankton and in in- 

 creased rates of anchovy filtering (MacCall 

 1980). In the course of switching to a preponder- 

 ance of feeding by filtering, the anchovy would 

 inadvertently filter anchovy eggs at a higher 

 rate. 



Egg production rate was sixfold higher in Pe- 

 ruvian anchovy compared with northern an- 

 chovy off California (Smith and Hewitt 1985a). 

 Of the 2,094 Peruvian anchovy eggs produced 

 per square meter, 67% died on the first day 

 while off California; of the 300 northern anchovy 

 eggs produced per m", 22% died on the first day. 

 While anchovy are omnivorous in both areas. 



phytophagy is considered more pronounced off 

 Peru (MacCall 1980). 



Pauly (1987) found that knowledge of the rate 

 of cannibalism appears crucial for managing the 

 Peruvian anchovy. The rate of cannibalism on 

 eggs is admittedly high, but it may be that the 

 variation in the rate of cannibahsm is insufficient 

 to explain the variation in the rates of recruit- 

 ment. It remains to be seen if interannual differ- 

 ences in the relative proportion of feeding by 

 filtration on the one hand and biting on the other 

 hand are sufficient to explain differences in re- 

 cruitment. 



The difference in conclusions regarding north- 

 ern anchovy egg cannibalism between Peterman 

 et al. (1988) for California anchovy and Pauly 

 (1987) for the Peruvian anchovy, may be due to a 

 simple difference in population density between 

 the two regions, California having a much lower 

 population density than Peru. Pauly (1987) 

 showed the relationship between population size 

 and egg mortahty: a better relationship would be 

 between local population density and egg mortal- 

 ity (Csirke 1980; MacCall 1980; Ware and Tsu- 

 kayama 1981). 



Egg mortality rate and adult biomass density 

 are closely related in interregional and interan- 

 nual egg production assessments (Fig. 3). The 

 relationship is surprisingly close when one con- 

 siders biomass of juvenile anchovy, the biomass 

 of other filtering fishes, particularly sardines, 

 and the biomass of invertebrate predators have 

 been neglected. In Table 6, one can see that 

 northern anchovy assessment studies off Cali- 

 fornia (Stauffer and Picquelle 1981; Stauffer and 

 Charter 1982; Picquelle and Hewitt 1983, 1984; 

 Hewitt 1985; Bindman 1986) and South Africa 

 (Armstrong et al. 1988) differ a gi-eat deal from 

 the assessment study reported in this paper for 

 Peru. The median biomass density of adult an- 

 chovy was 6 g/m^ off California between 1980 

 and 1985, while the median percent eggs dying 

 each day was 16%/d. In the south Benguela Cur- 

 rent, the mean biomass density of adult anchovy 

 was 12 g/m", and the mean percentage of eggs 

 dying per day was 22% (Armstrong et al. 1988). 

 The median regional estimate for 1981 of bio- 

 mass density of adult anchovy was 33 g/m^ off 

 Peru, and the egg mortality was 61% egg mortal- 

 ity per day. While these observations suggest 

 cannibalism as a cause, the mortality rate of eggs 

 may only be a consequence of generally higher 

 populations of a wide range of other vertebrate 

 and invertebrate predators in prime schooling 

 fish-feeding areas. 



504 



