Comparison of the Mortality Rates of Pacific Sardine, 



Sardinops sagax, and Peruvian anchovy, Engraulis 



ringenSy Eggs off Peru^ 



Paul E. Smith, Haydee Santander, and Juergen Alheit 



ABSTRACT: Egg production and mortality rates 

 were estimated for populations of Pacific sardine, 

 Sardinops sagax, and Peruvian anchovy, Engraulis 

 ringens, off north and central Peru for August to 

 September 1981. Peruvian anchovy spawned in the 

 entire inshore region, principally within 56 km (30 

 nmi) of the coast. The Pacific sardine spawned 

 extensively in the northern and southern inshore 

 regions but not in the central region. While spawn- 

 ing of the two species coincided at the regional 

 scale, at the sample scale, the occurrence of eggs of 

 the two species was statistically independent. The 

 egg production rate of Peruvian anchovy was about 

 double that of sardine. The mortality rate of Pa- 

 cific sardine eggs was much higher than that of 

 Peruvian anchovy. The egg mortality rates of Per- 

 uvian anchovy and Pacific sardine were unchanged 

 at stations where Pacific sardine and Peruvian 

 anchovy eggs occurred together, relative to where 

 the two species' eggs occurred separately. Results of 

 this study are used to evaluate cannibalism as a 

 population limiting mechanism. Cannibalism ac- 

 counts for about 30% of egg mortality in ancho\'y- 

 Because of the schooling habit of sardines and 

 anchovies, cannibalism, though large-scale, may 

 not vary enough with population size to explain 

 variations in recruitment. 



In their review of clupeoid biology, Blaxter 

 and Hunter (1982) emphasized the need to esti- 

 mate instantaneous rates as the "only way to 

 establish a satisfactory linkage between the fish 



'A Contribution of the Cooperative Peruvian-German 

 Fisheries Research Project (PROCOFA). 



Contribution No. 171 of the Alfred-Wegener-Institute for 

 Polar and Marine Research. 



Paul E. Smith, Southwest Fisheries Center La Jolla 



Laboratory, National Marine Fisheries Service, NOAA, 



P.O. Bo.x 271, La .Jolla, CA 92038. 



Haydee Santander (Institute del Mar del Peru, Apartado 22, 



Callao, Peru), managed the field and laboratory work and 



participated in the writing and editing of this manuscript 



until her death in March 1987. 



Juergen Alheit, Alfred-Wegener-Institute for Polar and 



Marine Research, Postfach 12 01 61, Columbusstrasse, 28.50 



Bremerhaven, Germany (FRG). 



Manuscript Accepted June 1989. 

 Fishery Bulletin, U.S. 87; 497-.508. 



and their environment." These instantaneous 

 rates include production of spawn (Lasker 1985) 

 and mortality rate of the eggs and larvae. Preda- 

 tion is thought to be an important cause of mor- 

 tahty, and one form of predation, cannibalism, 

 has been proposed as a mechanism for establish- 

 ing an upper limit to population density (Ricker 

 1954; Hunter and Kimbrell 1980; MacCall 1980; 

 Alheit 1987; Pauly and Soriano 1987) of filter- 

 feeding, pelagic spawning, schooling coastal 

 pelagic fishes. 



MacCall (1983) has described abundance as a 

 composite of population density and geographic 

 e.xtent: northern anchovy, Engraulis nionlax, 

 appears to have a central area where density- 

 dependent processes have their greatest effect 

 on local density and a peripheral area where den- 

 sity-dependent processes are less important. 

 Reductions of density in the central area result 

 in contractions of the population to the central 

 area for Pacific sardines, Sardinops sagax, 

 (Murphy 1966) and northern anchovy (MacCall 

 1980). Off California both sardines and anchovy 

 contract to the same area (Ahlstrom 1967). 



Microstructure in the sea plays an important 

 role in interaction of predator and prey (Lasker 

 and Smith 1977) at the embryonic and larval 

 stages (Smith 1973; Vlymen 1977; Theilacker 

 1987). There may also be rapid changes in the 

 microstructure due to dispersal (Smith 1973; 

 Smith and Hewitt 1985b). McGurk (1987) has 

 proposed that small-scale pattern may make 

 mortality rates proportional to patchiness rather 

 than inversely proportional as originally sug- 

 gested by Brock and Riffenburgh (1960). 



Subsequent variations in recruitment may be 

 caused by environmental influences on first- 

 feeding larvae (Hjort 1926; Lasker 1975, 1978, 

 1981; Peterman and Bradford 1987), and recruit- 

 ment is regulated by the interaction of gi'owth 

 and mortality rates of juveniles (Shepherd and 

 Gushing 1979; Smith 1985; Butler 1987). Peter- 

 man et al. (1988) demonstrated that in a selected 

 subset of years there was no relationship be- 



497 



