218 



Fishery Bulletin 93(2). 1995 



lated present fishing mortality rates, indicates that 

 summer flounder stocks are overexploited (NMFS, 

 1993). The decline in the natural fishery, together 

 with recent success in culturing other flatfish spe- 

 cies, such as the Japanese flounder, Paralichthys 

 olivaceus (Sproul and Tominaga, 1992), and the Eu- 

 ropean turbot, Scophthalmus maximus (Person-Le 

 Ruyet et al., 1991), stimulated interest in the develop- 

 ment of technology for the culture of summer flounder. 

 Basic information on the ability to distinguish 

 starving from feeding P. dentatus larvae and juve- 

 niles will be useful for studies of both natural and 

 cultured populations. Studies on the occurrence or 

 frequency of starvation in either field populations or 

 aquaculture operations must be preceded by an ex- 

 perimental study in which specific starvation indi- 

 cators are validated for fish of known nutritional his- 

 tory. Therefore, the aim of our research was to evalu- 

 ate and compare alternative criteria for assessing 

 starvation effects at several stages during the early 

 life history of P. dentatus. We characterize P. dentatus 

 larvae and recently metamorphosed juveniles subjected 

 to conditions of starvation or ad libitum feeding, using 

 biochemical, morphometric, and histological criteria. 



Materials and methods 



Adult broodstock P. dentatus were collected from 

 Narragansett Bay, Rhode Island, and Long Island 

 Sound, Connecticut, and were held in laboratory fa- 

 cilities. They were spawned after artificial induction 

 with repeated carp pituitary injections (2.5 mg/kg) 

 during 8 to 12 consecutive days (Smigielski, 1975). 

 The fertilized eggs were distributed in 38-L glass 

 aquaria covered with opaque black plastic. Each tank 

 was filled with UV-treated filtered (1 |im) Nar- 

 ragansett Bay seawater (adjusted to 34 ±\%c salin- 

 ity by brine addition). Antibiotic (200 mg erythro- 

 mycin activity dissolved in 23 liters of water) was 

 added at one time in each tank, and water changes 

 were performed every 2-3 days to maintain water 

 quality. During the first week, the alga, Tetraselmis 

 suecica, was added to the water. No artificial sub- 

 strate was added to the aquaria. Water temperature 

 was maintained at 19 ±1°C throughout the experi- 

 ment. Overhead illumination adjusted to a natural 

 photoperiod and aeration were provided. 



Hatching began 55 hours after fertilization. Dur- 

 ing the next 4 days the larval digestive system be- 

 came morphologically ready to process external food 

 at the time of mouth opening (Bisbal and Bengtson, 

 in press). Since yolk resorption and mouth opening 

 are almost simultaneous, flounder larvae were fed 

 daily on rotifers, Brachionus plicatilis, cultured on 



T. suecica (Lubzens, 1987) after day 4. Newly hatched 

 Reference Artemia III nauplii (Collins et al., 1991) 

 were offered for the first time 18 days later, and the 

 rotifer supply was progressively reduced. Settlement 

 to the bottom began on day 45 after hatching. 



Available literature on the early life stages offish 

 and previous direct observations on flounder cultures 

 directed our interest toward four developmental 

 stages (Al-Maghazachi and Gibson, 1984; Blaxter, 

 1988; Youson, 1988). The effects of starvation were 

 evaluated at day 6 (early food ingestion, yolk com- 

 pletely resorbed), day 16 (these larvae have positively 

 ingested and processed food at least once or else they 

 would have died within 10 days after hatching), day 

 33 (at the beginning of metamorphic eye migration), 

 and day 60 (bottom-dwelling juveniles have meta- 

 morphosed) after hatching. At these times, sub- 

 samples of the larvae pool were randomly placed in 

 one of two 5-L tanks (25 larvae/L): one (control group) 

 receiving food ad libitum (i.e. Brachionus or Artemia ); 

 the other (starved group) devoid of food. Although 

 the presence of food in the gut was not systemati- 

 cally recorded, the performance of feeding motions 

 and active swimming were visually confirmed on an 

 individual basis. An extra subsample was processed 

 as described below in order to establish the basal 

 levels of the several parameters measured prior to 

 the initiation of the imposed starvation (time 0). 

 Based on previous observations on the progression 

 of starvation at different age intervals and constant 

 visual monitoring of behavioral changes, each group 

 (control and starved) was sampled at least three 

 times, from the beginning of the experimental expo- 

 sure until the onset of mortality. A larva was consid- 

 ered dead if it did not respond to gentle probing with 

 a glass rod. If that was the case, the larva was cap- 

 tured and placed under the dissecting microscope for 

 a confirmation of its status. At each sampling time 

 the same protocol was followed: 6 to 10 individuals 

 from each group were sampled for histological analy- 

 sis, 10 for morphometric and dry weight measure- 

 ments, and 10 for biochemical analyses. 



Morphometric measurements consisted of stan- 

 dard length, eye diameter, head height, and the pec- 

 toral girdle angle as defined by Ehrlich et al. ( 1976). 

 Measurements of live larvae were taken under a dis- 

 secting microscope equipped with an ocular microme- 

 ter accurate to 0.7 um. Pectoral angles were traced 

 under a camera lucida and measured on a digitizing 

 pad. Each fish was then rinsed in deionized water 

 and placed in a 60°C oven until a constant dry weight 

 was obtained. Weight was measured, on either a Met- 

 tler AE 240 balance or a Cahn C-31 electrobalance. 



Samples for biochemical analysis were rinsed in 

 deionized water and individually preserved in Eppen- 



