FISHERY BULLETIN: VOL. 74, NO. 3 



DISCUSSION 



Significance of Feeding Incidence 



For the past half century, there has been a 

 discussion in progress concerning the significance 

 of feeding incidence. Lebour (1921) called atten- 

 tion to the low feeding incidence of young 

 clupeoids and attributed this to rapid digestion of 

 food in the larval intestine. She was soon chal- 

 lenged by Hardy (1924) who, after observing 

 herring larvae defecating after capture, assumed 

 the low value to be an artifact produced by most 

 larvae voiding their guts. The subject has attract- 

 ed increasing interest recently. June and Carlson 

 (1971) and Kjelson et al. (1975) observed older 

 larvae of the menhaden, Brevoortia tyrannus, 

 defecating after rough handling and fixation. 

 Anchovy larvae have been observed defecating 

 rotifers and Gymnodinium while being handled in 

 the laboratory (John Hunter pers. commun.). 

 Gymnodinium is eaten by E. mordax larvae in the 

 laboratory (Lasker et al. 1970) and probably so in 

 the ocean (Lasker 1975). Rotifers and the veligers 

 of various species of molluscs in combination with 

 Gymnodinium sustain anchovy larvae in the 

 laboratory up to about 25 days of age (Lasker et al. 

 1970; Theilacker and McMaster 1971). Blaxter 

 (1965), however, in attempts to assess the effect of 

 Formalin -^ on food retention of herring larvae was 

 able to demonstrate that only 10% of the larvae 

 empty their guts due to Formalin fixation. 

 Detwyler and Houde (1970) studying laboratory- 

 grown larvae of scaled sardine Harengula pen- 

 sacolae, and bay anchovy, Anchoa, mitchilli, found 

 almost all of even the first feeding stages con- 

 tained food after samples of them were taken 

 from the plankton rich rearing tank and preserved 

 in 5% Formalin. Feeding incidence of clupeoid 

 larvae captured in plankton nets has been posi- 

 tively correlated with the availability of food by 

 Pavlovskaia (1958), Nakai et al. (1966), Burdick 

 (1969), Nakai et al. (1969), Bainbridge and Forsyth 

 (1971), and Schnack (1974). Blaxter (1965) cited the 

 wide variation and observed feeding incidence in 

 the literature concerning herring larvae. I believe 

 that much of the confusion has resulted from 

 many authors failing to consider the time of day 

 when larvae were caught (Figure 2) or the age of 

 the larvae (Figure 4). When these variables are 



•'Reference to trade names does not imply- endorsement by the 

 National Marine Fisheries Service, NOAA. 



taken into account, a series of observations of 

 feeding incidence can reveal valuable insights into 

 the tropho-dynamics of larvae. Feeding incidence 

 must be viewed only as an indicator of feeding 

 success because of the errors likely to be produced 

 by defecation or to the difficulty in detecting soft 

 bodied items such as Gymnodinium. 



Comparison of the feeding incidence in four 

 species of Engraulis (Figure 8) shows an increase 

 in feeding incidence over larval lengths of 3 to 4 

 mm. Following this relatively high incidence at 4 

 mm, there is a drastic drop in this value until 

 lengths of about 7 or 8 mm are reached. The mean 

 feeding incidences for the four curves in this 

 length range are 7 times higher for the 4-mm than 

 for the 8-mm larvae. Feeding incidence remains 

 low but relatively constant over the length range 

 from 8 mm to about 14 mm at which point it begins 

 to increase steadily over the remainder of the 

 larval period. The value for the 20-mm length of E. 

 nngens is based on only 12 specimens and, there- 

 fore, is not as reliable as values for other lengths. 



The available data for sardine larvae suggest 

 the same U-shaped curve. When the values for the 

 sardine (Figure 7) are compared to Figure 8 it is 

 seen that feeding incidence in relation to size falls 

 roughly between E. ringens and E. anchoita, 

 except that the decrease at intermediate sizes is 

 not as precipitous. Yamashita (1955) reported the 

 following feeding incidence values for larval 

 Sardinops melanosticta: for about 14 mm = 8%, 21 

 to 30 mm = 56%, and 31 to 40 mm = 81% . The 

 upward trend of these data is similar to those of 

 larger anchovy larvae; however, the values are not 

 comparable because the time of day of sampling 

 was not reported. It seems significant that the 

 shape of the curves of the four anchovy species 

 (Figure 8) are so uniform in their relation to each 

 other. Engraulis ringens is considerably higher 

 than all others (except for the value at 20 mm). 

 This probably is related to the rich plankton 

 conditions of its habitat. 



Clupeoid larvae visually detect prey, approach 

 it, and then strike from a characteristic S-shaped 

 posture. Proficiency of capture increases with age 

 as observed in the laboratory for the larvae of 

 herring and pilchard (Blaxter and Staines 1971), 

 sardine (Schumann 1965), and anchovy (Hunter 

 1972). These investigators also noted that the 

 volume of water searched increases with larval 

 age. Feeding incidence should, therefore, increase 

 markedly with age. Why, then, does the observed 

 feeding incidence drop so drastically for anchovy 



524 



