Friedlander and Ziemann: Impact of hatchery releases on recruitment of Polydactylus sexfilis 



37 



reared fish were recovered from the leeward coast or 

 south shore areas. Highest proportions of hatchery-to-wild 

 fish were found in Kailua Bay (11%) and at Waimanalo 

 Beach (36%) during the 1998 season. For the Kahana Bay 

 site, around \2'7( of the Pacific threadfin recovered in the 

 1998 reward program were hatchery-reared. For the 1999 

 reward program, Kailua Bay continued to have the high- 

 est percentage of hatchery-reared Pacific threadfin, 40%. 

 Hatchery-reared Pacific threadfin were also recovered at 

 Waimanalo Beach (8.7%), Mokapu Peninsula (9.7%), and 

 Kaneohe Bay (11.2%); the latter two sites are north of the 

 Kailua Bay release site. 



Hatchery-reared Pacific threadfin released in Kahana 

 Bay were recovered as far north as Kawela Bay (25 km) 

 and south to Waimanalo Beach (33 km) (Fig. 3). Fish re- 

 leased in Kailua Bay were recovered 19 km to the south at 

 Sandy Beach and 42 km to the north at Kahuku (Fig. 4). 

 Fish were captured at significantly greater distances 

 from the release sites in 1999 (x=15.2 km, SD=10.2) 

 than in 1998 (.r=11.0 km, SD=6.8) (7=2044.0, P=0.025). 

 Fish released in Kahana Bay were recaptured at signifi- 

 cantly greater distances from the release site (.v=14.6 km. 

 SD=11.0) compared to the Kailua Bay releases (.v=9.6 km, 

 SD=6.3; T=6129.0, P<0.001). 



The proportion offish recaptured in the reward program 

 for 1998 and 1999 was lowest in the largest release size 

 class (115-130 mm) and greatest in the middle two release 

 size classes (85-100; 100-115 mm). Because the number 

 of fish released in each size class in 1997 was different, it 

 was necessary to compare the number of fish recovered in 

 each size class in relation to the number released in that 

 size class. Eight fish recaptured from the 1996 release 

 were not included in our analyses. For both years and both 



locations, the catch of hatchery-reared Pacific threadfin 

 divided by the adjusted release number was greatest in 

 the smallest size class (70-85 mm) (Table 4). 



Expected number of fish recaptured in each size class 

 was calculated by taking the total number of recaptured 

 fish for each year from each release site and multiplying it 

 by the proportion of fish released in that size class at that 

 site during that year (Table 4). There were no significant 

 differences in number of fish recaptured in each size class 

 in relation to expected recapture numbers for both years at 

 the Kahana Bay release site ( 1998: f'-^m, df=3, P>0.05; 

 1999: x-=^.2A, df=3, P>0.05) and for Kailua Bay (1998: 

 ^-=5.74, df=3, P>0.05; 1999: ;f-=7.14, df=3, P>0.05). 



Angling (pole and line fishing) accounted for the great- 

 est number of hatchery-reared and wild Pacific threadfin 

 acquired in the reward program but had the lowest CPUE 

 among gear types (Table 5 ). Over both survey years, for wild 

 and hatchery-reared fish combined, angling accounted for 

 63.0% of the total Pacific threadfin catch, followed by gill- 

 nets ( 19.2% ), thrownets ( 13.9% ), and surround nets (3.9% ). 



Length of Pacific threadfin caught varied among gear 

 types, between years, and between hatchery-reared and 

 wild Pacific threadfin. For 1998, mean size of Pacific 

 threadfin captured was significantly different among gear 

 types (P=6.378, df=3, 632, P<0.001) and between hatch- 

 ery-reared and wild fish (P=11.833, df=l, 632, P<0.001). 

 Gillnets tended to catch larger fish, and surround nets 

 captured the smallest fish (Tukey multiple comparison 

 test results — gillnets: 255.5 mm > angling: 235.1 mm > 

 thrownet: 214.0 mm > surround nets: 193.7 mm). Mean 

 length for hatchery-reared Pacific threadfin pooled over all 

 gear types was 207.0 mm (SD=28.6) and mean length for 

 wild Pacific threadfin was 242.2 mm (SD=52.7). 



