Jacobson et al.: A biomass-based assessment model for Engraulis mordax 



713 



1000 T 



1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 



FISHING SEASON 



Figure 1 



Indices of abundance for northern anchovy, Engraulis mordax (log scale). 

 The SPOTTER index is based on fish spotter data, HEP is the historical 

 egg production index, EPI is the new egg production index, SONAR is based 

 on sonar data, and DEP is daily egg production method estimates of spawn- 

 ing biomass. 



anchovy begin on 1 July, end on 30 June, and are 

 identified by the calendar year on 1 July. We assumed 

 that indices for northern anchovy measured abun- 

 dance during the peak spawning period which is 

 about 15 February. 



Indices of abundance used to estimate northern 

 anchovy biomass included 1) spawning biomass es- 

 timated by the daily egg production method (DEP, 

 Gunderson, 1993; formerly called the egg production 

 method or EPM, Lasker, 1985), 2) a historical egg 

 production index (HEP, Lo, 1985), 3) our new egg 

 production index (EPI), 4) relative biomass of 

 schooled fish estimated from fish spotter data (SPOT- 

 TER, Lo et al., 1992), and 5) relative biomass of 

 schooled fish estimated from California Department 

 of Fish and Game sonar data (SONAR, Mais, 1974; 

 Methot, 1989). Indices that measured relative abun- 

 dance of northern anchovy (Fig. 1) were not all posi- 

 tively correlated and none of the correlations were 

 statistically significant. Lack of significant correla- 

 tion may have been due to relatively few years of 

 data used to compute some correlations, to impre- 

 cise indices, and to differences among indices in area 

 surveyed. Of particular concern is the lack of signifi- 

 cant positive correlation between the EPI and SPOT- 

 TER indices which are the only relative abundance 

 data available for recent fishing seasons. 



Daily egg production method (DEP) estimates of 

 spawning biomass during the 1979 to 1984 seasons 



(Table 1, Fig. 1) measured spawning biomass in met- 

 ric tons (t), rather than in relative units, and are 

 relatively precise (coefficients of variation (CV) less 

 than 27%). Only six DEP observations were available, 

 but the DEP data were important because they helped 

 scale indices of relative abundance for northern anchovy 

 to units of absolute biomass (Bence et al., 1993). 



SONAR data did not include variances that were 

 required in our model. We used the standard devia- 

 tion (e=0.439) of log-scale residuals for SONAR val- 

 ues from a previous study 2 and the relationship 



CV 



1 



(1) 



to obtain a crude estimate (46%) of the arithmetic 

 scale coefficient of variation. 



The SPOTTER index used in this study (Table 1, 

 Fig. 1) was similar to the one developed by Lo et al. 

 ( 1992 ) except that April to March, rather than Janu- 

 ary to December, annual periods were used to aggre- 

 gate data. The SPOTTER index value for the 1963 

 fishing season was anomalously low ( Fig. 1 ). Follow- 

 ing Lo et al. ( 1992), we excluded the 1963 value be- 

 cause the data collection program was new in 1963 

 and the information may not have been reliable. 



New egg production index (EPI) 



Our new egg production index (EPI) measures egg 

 production by northern anchovy during the 1979 to 



