FISHERY BULLETIN; VOL, 77. NO, 2 



Table 2. — Results of chi-square test of the hypothesis (see text) that tempera- 

 ture-specific grovrth in the northern anchovy is the same m sea and laboratory 

 (A-Ei and the parameters of the Gompertz growth curve fit to the data in each 

 sample (F-I). A, sample size; B, surface temperature in degrees Celsius; C. 

 growth curve compared with data; D, probability that growth curve fits data; E. 

 sign of significant deviations l0.05 level i from laboratory curve; F and G, Gom- 

 pertz parameters, A and a; H, standard error of regression; 1. growth rate at 8 

 mm (in miUimeters/day). Last row shows results for combined data (samples 

 with *). 



faster than expected. The larvae collected in 

 March 1977 along the transect extending seaward 

 from San Diego (DI-D61 grew about as fast as 

 expected, given the surface temperature at which 

 the sample was collected. 



Since the data in 7 of 12 samples deviated sig- 

 nificantly from the predicted Gompertz growth 

 curve, we fit the Gompertz growth function 

 (Zweifel and Lasker 1976) 



log,, iSL) = log,, iSL^) + 



An 



(1 



■) (1) 



where SL = standard length in millimeters 

 t = number of increments = age in 

 days postyolk absorption 

 SL„ = initial size, fixed at 4.2 mm 

 A„, a = parameters to be estimated 



to each using a nonlinear least-squares method 

 (Conway et al. 1970). 



The parameters of the fitted curves (Table 2) 

 were used to calculate a linear approximation of 

 the growth rate of 7.5-8.5 mm anchovy larvae 

 (Figure 3). The range of growth rates among the 12 

 samples collected at 13°- 16.2" C was bounded by 

 the growth rates of anchovy larvae grown in the 

 laboratory at 14° and 17.5° C. Growth was very 

 similar in all but three samples (Bl, CI, C2) al- 

 though the temperatures associated with these 

 nine samples spanned a range of 2.5° C. The 

 growth rate of 8 mm larvae calculated from a 

 Gompertz curve fit to the combined data of the 

 nine similar samples was 0.37 mm day. The mean 



TEMPERATURE rc) 



Figure 3. — Relationship between growth rate imillimeters/day 

 at 8 mm) and temperature for northern anchovy larvae from the 

 field (bare symbolsl and in the laboratory i circled symbols I. Field 

 samples, this study, A1-D6; composite of nine field samples, Bl, 

 CI, and C2 not included. iJ; Hunter 1976, laboratory-reared 

 prey. H; Kramer and Zweifel 1970, wild plankton for food, K; 

 Lasker et al, 1970, dinoflagellates only, LD, dinoflagellates plus 

 veliger, LV; Sakagawa and Kimura 1976, S; unpubl. data avail- 

 able at Southwest Fisheries Center, low temperature experi- 

 ment, E, periodic starvation experiment, F. Curve was derived 

 from the model of Zweifel and Hunter (see text footnote 21; it was 

 not fit to the growth rate data presented here. 



temperature of these nine samples (weighted by 

 number of larvae) was 15.04' C. The model of 

 Zweifel and Hunter (see footnote 2) predicts that 8 

 mm larvae reared at 15° C would grow at 0.395 

 mm/day. 



418 



