KOSLOW; FEEDING SELECTIVITY OF NORTHERN ANCHOVY SCHOOLS 



250 350 450 550 650 750 

 BODY LENGTH (pm) 



Figure L — Fraction consumed of zooplankton prey items by 

 northern anchovy schools as a function of prey body size ( A-D). 

 Linear regressions were performed on all data sets, excluding 

 from the regression data points past the first size class at which 

 the school consumed >90% of prey. A curvilinear fit to the data 

 based on a regression using a loge-transformation of values on 

 the abscissa is shown where a curvilinear relationship provides 

 as good or better fit. The regressions in Panels A-C include the 

 data from each species found in the panel. In Panel D, the linear 

 and curvilinear regressions in the upper left are based on data 

 for Calanus pacificus alone; the solid lower regression is based 

 upon data for Sagitta spp. and Acartia tonsa; the dashed oblique 

 line is based upon the total data set for the sampling period. 

 Dashed horizontal lines represent the fraction consumed of the 

 total available zooplankton biomass. The significance level for 

 each regression (which is also the significance level for each 

 regression coefficient) is indicated. Panels: A and B — August 

 1975; C— October 1976; D— April 1976; E— March 1976; Ivlev's 

 Electivity Index (E) as a function of the pigmented body length of 

 Evadne spp. 



the largest zooplankters (Table 4, Figure 1). Over- 

 all, the northern anchovy schools consumed 35- 

 50% of the total available zooplankton in the areas 

 sampled (Figure 1). 



On the summer cruise (August 1975), a diverse 

 assemblage of small plankton was present. Prey 

 consumption appeared to be a function primarily 

 of their size (expressed as their weight in micro- 

 grams carbon) (Figure lA). A single regression, 

 whether linear or curvilinear, adequately de- 

 scribes the northern anchovy's feeding selectivity 

 for such morphologically dissimilar organisms as 



copepod nauplii, Euterpina acutifrons and Para- 

 calanus parvus; small calanoid copepodites, Acar- 

 tia tonsa and P. parvus; and small chaetognaths, 

 Sagitta spp. There is no indication of species 

 preference. Nor is there evidence that prey density 

 significantly influenced the rate at which they 

 were consumed. While the relative concentration 

 of similar-sized prey items varied widely, they 

 were consumed at equivalent rates [Table 4, 

 compare the density and consumption of copepod 

 nauplii and Sagitta spp. (0.5-1.0 mm), P. parvus 

 CIII-VI and Sagitta spp. (1.0-1.5 mm), and A. 

 tonsa CIII-VI and Sagitta spp. (1.5-2.0 mm)]. 



The data for the northern anchovy's feeding on 

 the larvacean, Oikopleura spp., from the August 

 cruise was not directly comparable to data for the 

 other species sampled at this time because their 

 body length cannot be converted to a carbon value 

 for the whole organism (including "house"). How- 

 ever, the northern anchovy's feeding on Oiko- 

 pleura appeared to be a linear function of prey size 

 (Figure IB). 



The autumn cruise (October 1976) was charac- 

 terized by a low density of zooplankton — the 

 standing crop was a factor of 3-4 less than that 

 encountered during the April 1976 and August 

 1975 cruises — entirely dominated by small zoo- 

 plankton as in the August 1975 cruise (Tables 1, 

 4). Again the northern anchovy's feeding selectiv- 

 ity was positively related to prey size (Figure IC). 

 A curvilinear relationship here provides a better 

 fit to the data. 



Only on the cruise of April 1976 were both large 

 and small zooplankton present; however, the zoo- 

 plankton density was comparable to that found on 

 the summer cruise (August 1975) (Table 1, com- 

 pare the range of the prey sizes; Table 5, compare 

 Figure ID with Figure lA, C). As in the summer 



Table 5. — Maximum prey size and its density, prey size at 100% 

 consumption by northern anchovy schools as predicted from 

 linear regressions, and slope of linear regressions of the school's 

 feeding selectivity of prey size. 



137 



