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



tween the abundance of northern anchovy larvae 

 off California at 20 days of age and the recruit- 

 ment at age 1. It had previously been shown 

 that, for a shorter time series in the same habi- 

 tat, Pacific sardine recruitment at two years old 

 varied 15-fold, while the egg abundance in the 

 year of origin for the same years varied only 3- 

 fold (Smith 1978). 



A major barrier to estimating egg mortality in 

 pelagic schooling fish is the tendency of these 

 fish to aggi'egate closely for external fertilization 

 of spawn (Leong in press), presumably to con- 

 serve sperm and ensure high fertilization rates. 

 The eggs are subsequently dispersed by turbu- 

 lent diffusion (Smith 197*3; Smith and Hewitt 

 1985a). Schooled spawning, with its spatially in- 

 tense pattern, imposes a high sample variance on 

 the younger eggs, and dispersal reduces the 

 sample variance with time (Smith and Hewitt 

 1985a). 



Precise estimation of survival and egg produc- 

 tion rates requires large numbers of samples. 

 The intensity of patchiness is indicated by the 

 maximum values for egg counts. Smith (1973) 

 observed that the concentration of Pacific sar- 

 dine eggs off California was a maximum of 3,100 

 eggs/nr, and Hunter (1980) observed that a con- 

 centration of 1 d old northern anchovy eggs was 

 4,600 per m^. Walsh et al. (1980) reported the 

 results of sample analyses of Peruvian anchovy 

 eggs along the Peruvian coast. P^i'om five sam- 

 ples at lat. 10°S and seven samples at 15°S, they 

 found a maximum value of 11,500 Peruvian an- 

 chovy eggs/m-, and they also found Peruvian 

 anchovy larvae at 100 and 200 per m^. Santander 

 et al. (1982) reported on Peruvian anchovy and 

 Pacific sardine off Peru. For Peruvian anchovy, 

 they found a maximum egg concentration of 

 107,376 eggs/m^ (all ages). This count was from 

 the largest of 911 ichthyoplankton tows with 

 Peruvian anchovy eggs out of 4,028 Hensen net 

 samples between 1966 and 1979. For Pacific sar- 

 dine off Peru, the maximum egg concentration 

 was 58,500 eggs/m^ (H. Santander^). This count 

 was from the same set of 4,028 tows. 



Pacific sardine spawning off Peru has been es- 

 sentially stable between 1978 and 1981 (San- 

 tander 1981). What are the implications for se- 

 lection of spawning sites for the two pelagic 

 spawners? Do the two species partition the en- 

 vironment into separate spawning areas or do 

 they concentrate on oceanic sites favorable to 



^H. Santander, Institute del Mar del Peru, Apartado 22, 

 Callao, Peru, unpubl. data. 



both? What are the relative rates of egg produc- 

 tion and mortality where they co-occur? 



In this paper we describe regional and inter- 

 specific differences in egg production and mor- 

 tahty off Peru. The Peruvian anchovy data used 

 for this paper have already been used for esti- 

 mating egg production of the Peruvian anchovy 

 population off north and central Peru (Santander 

 et al. 1984). The Pacific sardine egg data off Peru 

 are adequate for determining egg production per 

 unit area, but contemporaneous adult reproduc- 

 tion rates are lacking; thus, sardine adult bio- 

 mass has not been estimated. 



METHODS 



Cruise plan, sampling methods, laboratory 

 methods, and data analysis have been described 

 fully by Santander et al. (1984; an English ver- 

 sion can be obtained from the author J. Alheit). 

 Only a brief account of these methods is pre- 

 sented here. 



The objective of the investigation was to en- 

 compass the Peruvian anchovy and Pacific sar- 

 dine spawning gi'ounds off northern and central 

 Peru with an intensive gind of ichthyoplankton 

 stations. The survey extended from Pisco (lat. 

 14°S) in the south to Punta Falsa (6°S) in the 

 north. The background, a computer program for 

 distributing the stations, and other technical de- 

 tails of the cruise plan are described in Smith et 

 al. (1983). The sampling stations were distri- 

 buted on transects perpendicular to the coast- 

 line. The inshore spacing of transects was 10 

 nmi. Thirty-five transects extended 30 miles 

 offshore and 18 transects extended 90 miles off- 

 shore. On all transects the sampling stations 

 were three miles apart. The total number of sta- 

 tions was 925. The area under investigation was 

 subsequently divided into nine regions for some 

 aspects of the analysis (Fig. 1). 



Eggs were sampled with the CalVET net 

 (CalCOFI vertical egg tow; Smith et al. 1985). 

 Its mesh size was 333 jxm, and its mouth area 

 was 0.05 m^. The net was towed vertically from a 

 70 m depth to the surface within 1 minute. The 

 net filtered 3.5 m^ of water. 



Following sorting, the eggs were aged accord- 

 ing to their developmental stage, the surface 

 water temperature, and the time of day when 

 the sample was collected. The following is a brief 

 summary of the method for determining Peru- 

 vian anchovy egg production; a complete de- 

 scription of these techniques for northern an- 

 chovy is found in Picquelle and Stauffer (1985). 



498 



