FISHERY BULLETIN: VOL. 83. NO. 4 



ern anchovies caught by midwater trawl. The gear 

 and sampling procedures are described by Mais 

 (1974). The data set covers the period 1966-1983 and 

 consists of 101 cruises. Twenty-three cruises extend- 

 ed north of Point Conception, 77 cruises occurred 

 in southern California and northern Baja California, 

 and 8 cruises extended into southern Baja Califor- 

 nia. Several cruises extended into more than one 

 region. There were a total of 4,166 trawl hauls, of 

 which 3,017 contained anchovies. Standard lengths 

 were normally taken from about 25 anchovies in each 

 trawl haul in which they occurred; otoliths, for 

 aging, were usually taken from a subsample of up 

 to 10 fish. A total of 60,082 northern anchovies were 

 measured, of which 20,772 were aged by California 

 Department of Fish and Game personnel with 

 methods developed by Collins and Spratt (1969). 



For purposes of determining age composition and 

 growth rates of northern anchovies, it was assumed 

 that February was the birth month of all fish sam- 

 pled. Anchovies off California and Baja California 

 have a peak in spawning in February- March; however, 

 some spawning does occur all year (Ahlstrom 1966). 

 The age determinations used in this report are, of 

 course, not accurate to the month. They are based 

 on the number of annuli, the seasonality of annuli 

 formation, an assumed February birth month, and 

 the month the fish were caught. Annuli formation 

 occurs in May in California (Collins and Spratt 1969). 

 A 1-yr-old anchovy would therefore be an anchovy 

 caught in February with no annuli on its otoliths. 

 A lV2-yr-old anchovy would be a fish with one an- 

 nuli which was caught in August. An anchovy born 

 in February would be 15 mo old when its first an- 

 nuli was formed. An anchovy spawned in September 

 would be 8 mo old if it formed its first annuli in its 



first May and 20 mo old if it formed its first annuli 

 in its second May. 



Differences in growth between geographical 

 regions were compared by linear regression analyses 

 which included anchovies of 1 V2 yr of age and older. 

 Evaluation of the relationship between age composi- 

 tion and water depth in the area of capture was made 

 by grouping the samples into depth strata. The strata 

 were established partially on the number of 

 observations. 



lb determine growth rates, the data were process- 

 ed with a computer program (Mallicoate and Par- 

 rish 1981) which calculates and plots the mean 

 length, standard deviation of length, and length 

 range by age and month. The program also tabulated 

 the number of observations by age and month; these 

 data were used for determining age composition. 



Our preliminary analysis of the sea survey data 

 showed an alongshore cline in the mean length of 

 northern anchovies (Fig. 1). It also revealed a bias 

 in the selection of fish to be aged. Aged fish were 

 generally larger at all latitudes than were the un- 

 aged fish. This bias is apparently due to a consis- 

 tent tendency for samplers to pick larger anchovies 

 for the subsample which was aged. Anchovies < 100 

 mm SL were particularly susceptible to not being 

 selected for aging (Tkble 1). On checking with the 

 field biologists who took the data, we found a second 

 source of bias which occurred only in trawl hauls of 

 exclusively small fish. When trawl hauls were con- 

 sidered by the sampler to be "obviously" all young- 

 of-the-year fish, there was a tendency not to take 

 otoliths for age determination. These sampling 

 biases affect the analyses of age composition 

 presented in this report but do not affect the grow^th 

 analyses. 



Table 1.— Two types of length bias in sampling northern anchovies in the Sea 



Survey Program. 



'Bias 1 is the decision to sample for otoliths (i.e., the proportion of fish in trawl hauls which 

 were sampled for otoliths). 



^Bias 2 is the selection of larger fish by the sampler (i e., the proportion of fish in trawl hauls, 

 which were sampled for otoliths, for which otolith samples were taken). 



484 



