NOTE Booman et a!.: Responsiveness of starved Engraulis mordax larvae to attacks by adult anchovy 



709 



The histological condition of the liver proved 

 to be the best character for diagnosis of starva- 

 tion in 9 mm larvae. Liver condition has been 

 used for the diagnosis of starvation in fishes in 

 combination with other characters by O'Connell 

 (1976), Theilacker (1978), and Watanabe (1985), 

 and liver alone was used by Storch and Juario 

 (1983) and Storch et al. (1984). In the results sec- 

 tion, we relate histological condition of the liver 

 to days of starvation and responsiveness to 

 predatory attacks. 



Results 



Response and escape probabilities 



The proportion of larvae responding to attacks 

 of adult anchovy progressively decreased with 

 starvation. Forty percent of fed larvae responded 

 to the attacks of predators, about the same pro- 

 portion determined for 8.5 mm anchovy by Folk- 

 vord and Hunter (1986, test of binomial propor- 

 tions, P 0.64). After 4 days of starvation 23% 

 responded, and by 7 days of starvation 17% 

 responded (Fig. 1). The higher value for both fed 

 and starved larvae on day 1 could partly be due 

 to the predator inexperience in preying on an- 

 chovy larvae, since during the acclimation period 

 the predators had been fed only live adult 

 Artemia. 



Considerable variability existed in the percentage of 

 larvae that responded to predators in the fed groups 

 from day to day (range of daily averages 35-54%) and 

 among predator and prey groups (range of group aver- 

 ages 36-45%). Since a series of trials using fed larvae 

 always accompanied those with starved larvae, it was 

 possible to adjust for this interexperiment variability 

 by dividing the fraction of starved larvae that re- 

 sponded by the fraction of fed larvae that responded 

 during the same experiment. After one day of starva- 

 tion (that is, 24-30 hours after transfer) this ratio was 

 0.77, indicating that the response probability for larvae 

 deprived of food for 1 day was 23% lower than that 

 of fed larvae. The regression for these calculated ratios 

 indicates that larvae starved for 4 days were 40% less 

 likely to respond than fed larvae, and those starved for 

 7 days were 58% less likely to respond (Fig. 2). 



Very few larvae escaped predators regardless of 

 their nutritional state. Thus, it was not possible to 

 evaluate the proportions escaping as a function of the 

 duration of starvation. Considering all data combined, 

 starved larvae were less successful in escaping pred- 

 ators (0.8% vs. 3.4% fed larvae; test of binomial pro- 

 portions, P 0.027). Escape probabilities were lower in 

 fed larvae in this study (3.4%) than in Folkvord and 



3 4 5 



ELAPSED TIME (days) 



Figure 2 



Responsiveness to anchovy predators of starved northern anchovy 

 larvae (relative to fed larvae tested in the same experiment) shown 

 in relation to duration of the starvation period and the condition of 

 larval hepatic tissue. Data from days 6 and 7 were combined to ob- 

 tain sufficiently high numbers of starved larvae. Equation for the line 

 is Y = 0.824- 0.057 X (R 2 0.87, SE est. 0.05). Histological character- 

 istics of the liver of starved larvae are indicated with letters and 

 arrows. (A) Glycogen vacuoles in hepatocytes, (B) no glycogen, struc- 

 tures unaltered. (C) hepatocytes membranes damaged, dark nuclei, 

 and (D) cell atrophy, nuclei pycnotic and irregular. 



Hunter (1986, 12%) for 8.5 mm larvae (test of binomial 

 proportions, P 0.01). Some obvious differences existed 

 between the studies. In the present study, body size 

 of the anchovy predators was about 10% longer, tem- 

 perature in the experimental tank about 5°C lower, and 

 the test tank acclimation period of predator groups was 

 longer. Any of these factors could affect the probabil- 

 ity of escape. 



Relation between prey responsiveness 

 and histological condition 



Larvae deprived of food for 1 day had no vacuoles in 

 their liver, but cellular structures showed no alterations 

 (Fig. 3). The vacuoles seen in fed larvae presumably 

 contain glycogen (O'Connell and Paloma 1981). Disap- 

 pearance of vacuoles after 24 hours of food depriva- 

 tion has also been observed by Watanabe (1985) in 

 freshwater gobiid larvae reared at 15-20°C. At this 

 stage of starvation, anchovy larvae were 23% less 

 responsive to predators than were fed larvae. After 4 

 days of starvation, the hepatic tissue began to deteri- 

 orate; cell outlines were hardly distinguishable, and the 

 nuclei stained darkly. Larvae with these characteristics 

 were about 40% less responsive than were fed larvae. 



