458 



Fishery Bulletin 94(3), 1996 



Lagoon releases 

 Inlet releases 



10 20 30 40 



Weeks after first release 



Figure 2 



Release microhabitat effect on dispersal of released cul- 

 tured fish. Proportions (± SEM, n=6 lots) of cultured striped 

 mullet recaptured outside of the release habitat, Kahaluu 

 Stream, following downstream (inlet I releases at the shore 

 next to the stream mouth, and releases 300 m upstream 

 (lagoon). 



Release microhabitat effect on enhancement 

 impact Initial habitat selection was strongly af- 

 fected by release microhabitat. There was greater 

 dispersal away from the release site by fish released 

 next to the inlet of Kahaluu Stream than by fish re- 

 leased about 300 m upstream in the lagoon (Fig. 2.) 

 This pattern was similar following both spring and 

 summer releases (Table 4). Most fish released up- 

 stream remained in Kahaluu Stream throughout the 

 study, whereas most fish from the inlet releases 

 moved to other nursery sites in the bay. This differ- 

 ence in dispersal patterns was statistically signifi- 

 cant by the second collection date (P<0.003, n=6 re- 

 lease lots per treatment [3 spring + 3 summer]) and 

 was observed through mid February 1993 (P<0.01 

 in collections Aug, Oct, and Jan; P<0.05 in Nov, Dec, 

 and Feb; not significant in Jun, Sep, Mar, and Apr). 

 Recapture rates and growth rates were unaffected 

 by release microhabitat. Growth curves for fish re- 

 leased at the inlet and lagoon were intermingled (Fig. 

 3) as were plots of numerical abundances of cultured 

 fish in collections (Fig. 4). Statistical comparisons 

 were nonsignificant («=6, P>0.08) for all collection 

 dates. 



SAR and release-season effects on enhancement 

 impact 



Overall SAR effect on recapture rates Fish size- 

 at-release affected survival of mullet released in this 

 study. The smallest size group (45-60 mm TL) had 

 significantly reduced survival in comparison with 

 other sizes (Fig. 5 ). Fish <60 mm TL were recaptured 



Lagoon releases 

 inlet releases 



10 20 30 40 



Weeks after first release 



Figure 3 



Mean length (TL [+SEM, n=6]) of cultured striped mullet 

 recovered in net samples made over the course of the study, 

 for individuals that were 70-85 mm TL when released. 



Lagoon releases 

 Inlet releases 



f 







10 20 30 



Weeks after first release 



Figure 4 



Catch per unit of effort (CPUE) for cultured striped mullet 

 from collections made at four nursery habitats (combined 

 data) in Kaneohe Bay, including the release site. Total 

 number of tagged fish recaptured in 120 cast-net samples 

 taken monthly (30 samples per month at each nursery habi- 

 tat) compared with those taken over the 11-mo study period. 



at less than half the frequency of larger fish released 

 (Fig. 5; P<0.001; P<0.002 in orthogonal contrast in a 

 comparison of SAR-1 with SAR1-2, -3, and -4 com- 

 bined; n =6 with recapture data from spring and sum- 

 mer releases combined ). For the four SAR groups >60 

 mm TL (in the aggregate data set, spring and sum- 

 mer releases combined), there were no significant 

 differences among recapture frequencies (P>0.30 

 for any comparison, spring and summer releases 

 combined). 



