FISHERY BULLETIN: VOL. 75, NO. 3 



Atlantic herring, C. harengus, from the Gulf of 

 Finland reported that, at observed plankton con- 

 centrations in the field, the calculated time of 

 feeding was 15 h. This coincided exactly with the 

 length of day. Laurence (1971a), working with 

 the stipulation of a 14-h feeding period for large- 

 mouth bass larvae, found that prey concentrations 

 of7.0cal/liter (400 organisms/liter) were limiting. 

 The results of this research show that simulated 

 critical prey densities, below which winter floun- 

 der larvae do not have enough daylight hours for 

 feeding to meet growth and metabolic energy 

 requirements, actually vary with age and stage 

 of development (Figure 7). The critical densities 

 range from a high of 5.7 (0.8 nauplius/ml) to a 

 low of 2.1 cal/liter (0.3 nauplius/ml) when feeding 

 behavior has been established but before growth 

 and metabolic demands are high. Critical density 

 then increases until initiation of metamorphosis 

 when it remains fairly constant around 4.5 cal/ 

 liter (0.6 nauplius/ml). Results such as these have 

 not been quantitatively reported in the literature 

 before. Most previous laboratory studies for a 

 variety of species delineate constant critical prey 

 densities for the larval period usually in the range 

 0.1-1.0 organism/ml (Kramer and Zweifel 1970; 

 O'Connell and Raymond 1970; Saksena and 

 Houde 1972; Laurence 1974; Houde 1975), al- 

 though Rosenthal and Hempel (1970) reported 

 that prey densities for optimum feeding (not crit- 

 ical densities) for larval Atlantic herring were 

 higher for younger than older larvae. 



The critical prey densities for larval survival of 

 approximately 0.5 organism/ml noted in this and 

 the other cited laboratory research are somewhat 

 disparate with densities described from field data. 

 Lisivnenko ( 1961 ) noted that larval Baltic herring 

 were much less abundant in years when prey 

 abundance was <0.01 organism/ml. Sysoeva and 

 Degtereva (1965) reported that the minimum 

 abundance of Calanus finmarchicus, when the 

 intensity of feeding of cod, Gadus morhua, larvae 

 decreased, was from 0.01 to 0.005/ml and that a 

 concentration of 0.02/ml provided sufficient food 

 for survival. It is my opinion that the results re- 

 ported for laboratory studies may be more accu- 

 rate than the field study data presented thus far. 

 The laboratory studies represent highly con- 

 trolled experiments with accurate counts of prey 

 organisms. On the other hand, the field studies 

 give estimates of prey abundance which represent 

 average densities over linear or oblique sampling 

 distances. Planktonic prey organisms have conta- 



gious distributions and larvae may well be associ- 

 ated with "patches" of prey that are more densely 

 concentrated than indicated by plankton net tows 

 (Wyatt 1973). Many larval fish researchers feel 

 that density dependent mechanisms control 

 larval survival (Cushing and Harris 1973), and 

 the concept of contagious distributions in which 

 larvae and prey are associated in "clumps" that 

 may or may not be associated and occupying the 

 same area is one of the most logical ways to ex- 

 plain the fluctuations noted for natural larval 

 mortality. Also, field zooplankton sampling de- 

 signs rarely use nets with mesh smaller than 

 200 /xm. Most of the significant food organisms 

 utilized by larval fishes especially in the early 

 stages are <200 /xm in smallest dimension (Houde 

 1973) and would be lost in field sample estimates. 

 Use of the plankton pump may prove to be more 

 accurate in locating patches of zooplankton and 

 sampling the size organisms that larval fish con- 

 sume. Recently, Heinle and Flemmer (1975), 

 using a moving plankton pump, reported concen- 

 trations of nauplii of Eurytemora affinis in the 

 Chesapeake Bay area as high as 2.8/ml with con- 

 centrations of 1.0-1.8/ml not at all uncommon. 

 These concentrations are more than adequate for 

 good growth and survival of winter flounder lar- 

 vae and many other larval species. 

 . The initial, predicted decrease in metabolic 

 energy expended (Figure 8) during the period of 

 feeding initiation and shortly after ( 10-30 /xg dry 

 weight) is undoubtedly explained by the increased 

 feeding success with experience by first feeding 

 larvae. First feeding individuals have a lower 

 success ratio of captures and have to expend more 

 energy in searching for prey than older and more 

 accomplished feeders. This success or fail period 

 is critical to eventual survival and is relatively 

 short in duration for winter flounder, occurring 

 during the first 8 days after feeding begins at 

 8°C. The increase in metabolic energy expended 

 from 30- to 500-/xg dry weight after successful 

 feeding establishment is due to normal increases 

 in energy demand for all processes with rapid 

 increases in size usually seen in larval fishes. The 

 leveling off of metabolic energy demand during 

 the metamorphosis period (500-1,000 /xg dry 

 weight) may be unique to flatfishes due to marked 

 morphological and behavioral changes (Laurence 

 1975) and increased predatory efficiency requir- 

 ing less energy expenditure. 



The decrease in metabolic energy expenditure 

 with increasing prey concentration is logically 



540 



