hatch immediately upon collection but most 

 hatched 2 and 3 d later. 



Thermal tolerance experiments were performed 

 using a thermal gradient apparatus. Details of the 

 design and the operating characteristics are given 

 in Barker and Stewart (1978). In each experiment 

 groups of 10-15 larvae were transferred from the 

 ambient temperature water to test tubes contain- 

 ing 25 ml of water at the test temperatures in the 

 thermal gradient apparatus and were exposed for 

 5-60 min. After exposure the larvae were returned 

 to the ambient water. The addition of the larvae to 

 the apparatus also involved the addition of about 5 

 ml of ambient temperature water which resulted 

 in a slight lowering of the test temperatures. 

 About 10-15 min were needed for the test tempera- 

 tures to be reached again, so the temperatures 

 given in the results section for the 5 min exposures 

 were those reached after 5 min. 



A control was maintained outside of the thermal 

 gradient apparatus for all experiments. These lar- 

 vae were held at ambient temperatures and were 

 subjected to the same handling procedures as 

 those larvae which received the exposures to ele- 

 vated temperatures. 



All larvae were observed for mortality 2 h after 

 testing, at which time any dead larvae were re- 

 moved. The remaining larvae were observed again 

 after 24 h. Dead larvae could usually be recognized 

 by their opaque appearance and any larvae that 

 showed no movement when prodded were also con- 

 sidered dead. 



Each experiment was replicated once for the 

 smooth flounder larvae and the Atlantic herring 

 larvae. The replicate experiments were performed 

 for the rainbow smelt larvae in brook water and 

 also in seawater (28.8 "L , after acclimation for 1 d at 

 13°±1° C). All larvae tested were <5 d old and had 

 yolk sacs. 



Results 



The results of all experiments are presented in 

 Table 1. The Atlantic herring larvae were exposed 

 to AT"s ranging from 16° to 25 ° C for exposures of 5 , 

 15, 30, and 60 min. It appears that the larvae 

 acclimated to 8° C survived AT's up to 17° C for up 

 to 60 min. The larvae survived higher tempera- 

 tures at shorter exposure times. In all cases, ex- 

 cept the 5 min exposures, the mortalities in- 

 creased from approximately to 100% over a range 

 of2°C. 



The smooth flounder larvae were exposed to 



Table l. — Percentage mortalities of larvae of Atlantic herring, 

 Clupea h. harengus, smooth flounder, Liopsetta putnami, and 

 rainbow smelt, Osmerus mordax, in fresh and saltwater, under 

 different time-temperature combinations. Values for each of the 

 replicate experiments are listed. The temperature values in 

 parentheses indicate the adjusted temperatures for the 5 min 

 exposures as explained in the text. 



'Control (ambient temperatures). 



AT's ranging from 20.0° C to 30.2° C for exposures 

 of 5, 30, and 60 min. The larvae acclimated to 4° C 

 survived AT's up to 21.4° C for 60 min with negli- 

 gible mortality. The mortalities increased from ap- 

 proximately to 100% over a range of about 5° C in 

 the 30-60 min exposures. At the 5 min exposure 

 mortality increased from 6.7% (considered 

 background mortality) to 100% over a 8.7° C range 

 (25.2°-33.9°C). 



The rainbow smelt larvae held in freshwater 

 were exposed to AT's ranging from 11.3° to 19.4° C 

 for exposure times of 5, 30, and 60 min. The larvae 

 survived a AT of 13.6° C for up to 60 min, and at 

 exposures of shorter duration the larvae survived 

 higher exposure temperatures. For exposures of 

 30-60 min the mortalities increased from to 

 100% over a range of 4.0° C (26.8°-30.8° C). This 

 range was 1.8° C for the 5 min duration exposure. 



The rainbow smelt larvae which had been ac- 

 climated in seawater at 13 ° ± 1 ° C , after hatching in 

 freshwater, were exposed to AT's ranging from 



199 



