TEMPERATURE SELECTION BY YOUNG TOPSMELT 525 



structure. Isotherms were shifted during an experiment, and hot and 

 cold ends were reversed between experiments in an attempt to 

 partition position from temperature preferences. 



Topsmelt are gi'egai'ious, and we tested them in groups of six to 

 nine fish to assess the effects of temperature on populations rather 

 than on unnaturally isolated individuals. The fish were placed in the 

 experimental apparatus when the water was uniformly at their 

 acclimation temperature. A gradient was established about the fish 

 after they were habituated to the new surroundings. Larvae were 

 observed every 5 min for 4 to 5 hr and juveniles every 15 min for 6 

 to 8 hr or until a constant response was attained. Ehrlich et al. 

 (1978) suggested that the time required to reach the final 

 preferendum is related to the difference between it and the 

 acclimation temperature. In all 16 species of temperate marine fishes 

 tested, the final preferendum was attained within the experimental 

 times used (Ehrlich et al., 1978). 



The number of fish observations was defined as the total number 

 of observations of the experimental group times the number of fish 

 in the group. The frequency of occurrence of all experimental 

 temperatures was not uniform because of the shifting of the gradient. 

 This caused a bias in the number of fish observed at a particular 

 temperature when numbers were summed over an entire experiment. 

 To compensate for this, we adjusted the data before calculating the 

 mean and modal selected temperatures by using the number of fish 

 per total occurrence of a particular temperature rather than the 

 actual number of fish at each temperature (see Ehrlich et al., 1978). 



Field Investigation 



Plankton samples were collected at the southern end of San 

 Diego Bay, San Diego, Cahf., on approximately a 3- week basis from 

 June 1972 through June 1973, to relate the distribution of atherine 

 larvae to field temperatures. A 0.5-m-diameter plankton net with 

 0,505-mm mesh openings and a centrally located flowmeter was used 

 for collections, which were made during the day. The tows were 

 made just below the surface for 2 min at a speed of ~4 knots. 

 Specimens were preserved in 4% Formalin. 



RESULTS AND DISCUSSION 



Juvenile topsmelt {Atherinops affinis) swam freely along the 

 length of the experimental chambers before the temperature gradient 

 was established but showed some attraction to both ends. A total of 

 53 fish were observed during seven experiments; they were distrib- 



