WEINSTEIN ET AL.: RETENTION OF THREE TAXA OF POSTLARVAL FISHES 



the largest mean catch. This pattern paralleled 

 the flow of river water between tides; i.e., water 

 tended to move upriver on the eastern shoal and 

 returned on the west. If this phenomenon is real, it 

 indicates the existence of a large-scale circulation 

 pattern for postlarval populations. 



The tide by depth interaction clearly described 

 the immediate relationship of these organisms to 

 tidal flows. Whereas the depth distribution of spot 

 on flood tides was fairly uniform, bottom and mid- 

 depth concentrations on ebb often exceeded those 

 of surface values. Results were not quite as clear 

 for Atlantic croaker because of their general bot- 

 tom orientation; nevertheless, a tidal response 

 was still evident for this species (Figure 4). 

 Paradoxically, flounder showed little response to 

 tide (compared with the main effect result), al- 

 though a pattern similar to that of the other 

 species is shown in the mean concentrations in 

 Figure 5. 



All of these comparisons are potentially 

 influenced by diel activity, i.e., by downward mi- 

 gration during the day. Mean bottom values are 

 influenced by this effect on both flood and ebb 

 during daylight hours. One way of isolating the 

 effect of photoperiod would be to examine the in- 

 teraction of the three main effects (Table 3). 

 Unfortunately, this interaction was rarely sig- 

 nificant. Lack of significance may be a conse- 

 quence of the use of a logarithmic scale in making 



comparisons. Also, the power of tests on this 

 three-way interaction is considerably less than 

 that of tests of main effects and of two-way interac- 

 tions. That diel migration was not entirely respon- 

 sible for the observed patterns may be seen in the 

 overall (24 h) differences between flood and ebb. 

 Since two flood and ebb tides were sampled over 

 each 24-h period, the effect of diel activity should 

 be manifested on both tides; i.e., bottom orienta- 

 tion should occur on flood as well. Table 3 indicates 

 that this was not the case. Furthermore, a perusal 

 of the individual strata in Figures 3-5; e.g., an 

 examination of surface night concentrations 

 alone, shows that a clear tidal response was exhib- 

 ited by all three species. 



Length-Frequency Distributions 



The possibility that buoy 50 was located within 

 a primary nursery zone was alluded to earlier. 

 This contention is also supported by length- 

 frequency data which show that larger (older) fish 

 tended to accumulate upriver near buoy 50. Un- 

 fortunately, larger fishes were probably not cap- 

 tured quantitatively since gear efficiency drops off 

 rapidly after about 30 mm SL (Copeland et al. see 

 footnote 8). Hence, only a qualitative picture of 

 the age composition of a year class is possible. 

 Nevertheless, distinct size differences occurred 

 between buoys as indicated in Figures 6 and 7. 



10 



ro 



Q 



< 



_J 



cr 



LU 

 CD 



5 01 



001 



14-15 MAR 78 



BUOY 32, n = 2687 

 BU0Y50,n=0 



6 10 14 18 22 26 30 34 



LENGTH (mm) 



• • BUOY 32, n- 2287 



^— -o BUOY 50, n= 3250 

 1 1 -12 APR 78 



6 10 14 18 22 26 30 34 



LENGTH (mm) 



6 10 14 18 22 26 30 34 



LENGTH (mm) 



Figure 6. — Length-frequency distribution for spot, Leiostomus xanthurus, on the three collecting dates. This species was entirely 



absent from the vicinity of buoy 50 on 14-15 March. 



429 



