MAJOR: ASPECTS OF ECOLOGY OF STRIPED MULLET 



depth distribution were made to within 5-cm in- 

 tervals; the means given in the graphs being the 

 mean of a given interval (e.g., 47.5 cm for 45-50 

 cm). 



Statistical comparisons for depth or tempera- 

 ture distributions were significant (analysis of 

 variance, Ps^O.OOl) in all but a few cases. The 

 following comparisons were made, the exceptions 

 to P ^0.001 values being noted in parentheses: 1) 

 between observation intervals for given fish size 

 ranges and test salinity (no exceptions), 2) be- 

 tween salinities for given fish size ranges and ob- 

 servation interval (depth distribution for fish ^50 

 mm SL in observation interval 5 and salinities 15 

 and 34%o (P = 0.27); depth distribution for fish 

 20-30 mm SL in observation interval 1 and 

 salinities and 15%o (P <0.003); temperature dis- 

 tribution for fish 2=50 mm SL in observation inter- 

 val 1 and salinities and 15%o (P <0.002)), and 3) 

 between size ranges for a given test salinity and 

 observation interval (depth distribution for a sa- 

 linity of 34%o in observation interval 5 and fish size 

 ranges 30-50 and ^50 mm SL (P <0.004); temper- 

 ature distribution for salinity 34%o in observation 

 interval 5 and fish size ranges 20-30 and 30-50 mm 

 SL (P = 0.46)). 



Changes in depth distribution offish are readily 

 discernable in the histograms in Figure 3. At each 

 test salinity, with the exception of 30-50 mm SL 

 fish at 0%o salinity, mullet «50 mm SL moved 

 downwards in the tank to a mean depth of 47.5 or 

 52.5 cm by the last observation interval. Fish ^50 

 mm SL moved up to mean depths of 32.5 to 42.5 

 cm. Fish 30-50 mm SL at 0%o moved from an initial 

 distribution in the bottom half of the tank to a 

 mean of 52.5 cm during the remainder of the ex- 

 periments (observation intervals). As test 

 salinities decreased so did the final depth distribu- 

 tion for given fish size ranges, except for 30- to 

 50-mm SL fish. 



Changes in fish depth distribution were directly 

 related to tank temperature, since water tempera- 

 ture decreased with depth. However, temperature 

 values changed rapidly between depths 60 and 

 20-30 cm and were relatively isothermal and cold 

 between 20-30 cm and the bottom of the tank, and 

 isothermal and hot above 60 cm. As a result only 

 small differences in final depth distribution cor- 

 responded with relatively large differences in final 

 temperature distribution. 



The mean selected temperature tended to in- 

 crease between observation intervals 1 and 5 



at each test salinity for given fish size ranges. The 

 exception to this tendency was fish ^50 mm SL at 

 0%o salinity (20.9°-19.5°C). Fish «50 mm SL 

 (20-30 and 30-50 mm) tended to select higher final 

 observation interval 5) mean temperatures 

 (30.0°-32.3°C) than did fish ^50 mm SL (20.0°- 

 19.5°C) at each test salinity. For all fish size 

 ranges the final mean selected temperature tend- 

 ed to decrease as the test salinity decreased. This 

 decrease was greatest for fish >50 mm SL (29.0°- 

 19.5°C). 



The depth and temperature distribution results 

 taken together indicate that temperature selec- 

 tion was the more important, depth distribution 

 being secondarily related. Other gradients such as 

 light, pressure, and oxygen, if present, did not 

 appear to influence the distribution of the mullet 

 at least to the extent that temperature did. Fish 

 ^50 mm SL appeared to have a predilection to- 

 wards the surface whereas fish ^50 mm SL ap- 

 peared to be predisposed towards the bottom. This 

 was evident before the initiation of heating and 

 chilling in each experiment and during constant 

 temperature control experiments. As the experi- 

 ments progressed it also appeared as if mullet ^50 

 mm SL, in most instances, were 'Torced" down- 

 wards by rising temperatures. Similarly, mullet 

 ^50 mm SL were "forced" upwards by decreasing 

 water temperature, and then downwards by rising 

 temperature, such that their final temperature 

 and depth distributions were somewhat lower 

 than those for fish «50 mm SL in similar condi- 

 tions. 



Just how important the actual temperature and 

 depth distribution values selected by mullet are is 

 unknown. What does appear to be important is the 

 relative difference between distributions for fish 

 generally <50 mm SL as compared with those for 

 fish generally >50 mm SL at each salinity, and the 

 relative changes which occurred between salini- 

 ties for each fish length interval. The predisposi- 

 tion offish <50 mm SL towards the surface (and 

 higher temperatures) may be adaptive in the field. 

 Warm water rises and the warmest (hottest) water 

 is usually on the surface. By moving in the surface 

 layer fish <50 mm SL may be able to orient to- 

 wards the shallowest water inshore, which at low 

 tide should also be the warmest. The predisposi- 

 tion offish >50 mm SL towards the bottom of the 

 tank (and cooler temperatures) may similarly be 

 adaptive in the field. In this case movement away 

 from areas subjected to tide pool formation may be 

 important for survival. 



303 



