FISHERY BULLETDS": VOL 76. NO 4 



Th 



• ered or raised with the 



water les ei so cnai tney entered the water to a 

 - ---- - 2 -— Ge-r:e seration. frr-^ ''^ r "-:ttom of 



^_... . ::r.= -:. i^rther el.n:. vertical 



stratir. This edso aided removal of any 



supersaturated gases that Gift ' 1977 • reported to 

 '- r, 'tential problems in thermal gradients. A 

 nvion screen prever*" "- re=t organisms from 

 entering the area wr... _ ...r r.eating and cooling 

 took place and also provided a flat surface over the 

 hot: ' -. We subdivided the experimental 



gradient that consisted of 11 compartments each 

 containing a centrally located thermistor probe 

 into 2 1 NTsual units that were not visible to the test 

 organisms, by forming an additional compartment 

 centered midway between two thermistors. The 

 mean temperature of two adjacent probes was 

 used for these additional compartments. Day- 

 light-simulating fluorescent fixtures 'Dtiro-test 

 Vitalities', with the use of difFusers. pro\ided a 

 light intensity" of 60-70 fc at the water stirface. 



Gradient for Larval Fishes 



two nylon screens 1.5 m apart that defined the 

 experimental chamber. Sea water velocities of 0.1 

 to 0.9 mm s were selected based on known lars-al 

 swimming speeds Blaxter 1969: Rosenthal and 

 Hempel 1970: Hunter 1972'. because velocities in 

 this range woiild not present the larvae with any 

 significant difficult]." in maintaining a chosen posi- 

 tion. Wilson '1974' successfully employed veloci- 

 ties of up to 10.8 mms with pelagic marine fish 

 larvae in stud}-ing behavioral responses of pollu- 

 tants. 



Hunter and Thomas '1974^ showed that lan'al 

 ancho%'ies aggregated at patches of food ' Gym- 

 nodinium splendens i. All entering seawater was 

 filtered to 5 /im to rule out possible position selec- 

 tion by larvae based on presence of prey or- 

 ganisms. Eleven evenly spaced thermistor probes 

 coupled to a telethermometer were used to moni- 

 tor temperature. We 5ubdi\-ided the experimental 

 chamber into 21 \"isual compartments and en- 

 closed the trough in a lightproof box. A daylight- 

 simulating fluorescent lamp, with the tise of a 

 difFuser. produced 10-15 fc at the water STirface. 



The temperature gradient for larsal fishes < Fig- 

 ure 2) op>erated on a counter-current principle. 

 Alternation of hot and cold water entering the 

 experimental chamber between replicate runs 

 eliminated any potential rheotactic interference. 

 The inner experimental trough ' 1.75 m long x 5 

 cm diameten 0.8 mm wall thickness) contained 



DEFINITIONS 



The term "preferred temperature" has been 

 used in various contexts in the literature 'e.g.. 

 Brett 1952: Javaid and Anderson 1967: McCauley 

 and Tait 1970: Tatyankin 1972; McCauley and 

 Read 1973 ». Much of the variation in the use of this 



T0= /EW * 



h ®6 ^ 



 7- V ^^~^ 



V 



csOST VIEW WITH COVER 



*'^ri_ 



®-^)LJ-^^ 



# 





i- 





^ 



Figure 2. — &nall experimental dbamber for temperature selection measurements in larval fishes: aj experimental chamber, b) water 

 jaiiet, c) air line, d) drains, ei seawater input line, f > freshwater input line, gi daylight-simulating fluorescent light, h; light diffuser, i; 

 viewing slits, jj thermistor probes, ki door on lightproof cabinet, 1> supports for water jacket, m; water flow control valves, n> nylon 

 screen on ends of e^)erimental chamber. 



840 



