272 
Fishery Bulletin 95(2), 1997 
about 10 and 27 individuals per release lot through- 
out 36 weeks (Fig.2). However, as shown below, re- 
capture rate was in fact dependent upon the interac- 
tive effect of release season and size-at-release. (Note: 
for evaluating the effect of release-season, data can 
be compared only through weeks 35 or 36 following 
releases, the length of time fish were monitored af- 
ter summer releases; by the end of the study, fish 
released in the spring had been in the wild for an 
average of 45 weeks, 10 weeks longer than those re- 
leased in summer.) 
Dispersal patterns There were no clear seasonal 
trends in dispersal patterns. Cultured striped mul- 
let showed a strong tendency to remain in the vicin- 
ity of release sites, regardless of release season or 
size-at-release. Few of the 2,511 tagged fish recov- 
ered in samples had moved into other nursery habi- 
tats in the bay. The only significant movements ob- 
served were from release habitats into the streams 
located immediately to the north of each release site 
(Table 4). This pattern was repeated after spring and 
summer releases. There were isolated cases of fish 
moving from one release habitat to the other, as well 
as movement from release habitats into other nurs- 
ery habitats in the bay. But the magnitude of dis- 
persal out of release habitats and beyond the streams 
immediately north of those sites was negligible. Over- 
all, 90.8% ± 3.1% (SE) of the cultured fish collected 
through 36 weeks in the wild were recovered at the 
nursery habitats into which they had been released. 
Growth Growth after spring releases was similar 
to growth following summer releases. Length in- 
crease following releases is plotted in Figure 3 for 
fish from the 70-85 mm treatment group, which was 
representative of all 5 size-at-release groups. There 
was little change in mean length during winter 
months (from September 1991 through February 1992; 
weeks 20^45 following spring releases in Fig. 3). 
Release season effect on recapture frequencies 
among size-at-release groups Recapture frequen- 
cies ([number recaptured / number released] x 100% ) 
within size-at-release intervals revealed an obvious 
Figure 2 
Mean number of tagged cultured fish in samples following 
spring and summer releases into Kaneohe Bay. Data are 
means per release lot (± standard error [SE]; n = 6 lots per 
season [3 at each release site]). 
Table 4 
Movement patterns following 1991 releases in Kaneohe Bay. Release season and release site are identified for tagged fish recov- 
ered at the various collection (recapture) sites throughout the Bay. Recapture sites (and distances travelled) are ordered geo- 
graphically within collection dates, from the northernmost site (Waiahole Stream) to the southernmost site (Kaneohe Stream) at 
which tagged fish were collected (see Fig. 1). Totals for spring releases represent those through week 36; totals for summer 
releases are those through week 35. To compare results between release seasons over a similar time frame, data are excluded for 
weeks 40 and 45 after spring releases. 
Release season 
and recapture 
site 
Kahaluu Stream 
Kaneohe Stream 
Release season 
and recapture 
site 
Kahaluu Stream 
Kaneohe Stream 
n 
Distance 
(km) 
n 
Distance 
(km) 
n 
Distance 
(km) 
n 
Distance 
(km) 
Spring release 
Summer release 
Waiahole 
i 
3.05 
0 
15.00 
Waiahole 
0 
3.05 
0 
15.00 
Kaalaea 
92 
0.98 
0 
12.59 
Kaalaea 
14 
0.98 
0 
12.59 
Kahaluu 
509 
0 
1 
12.04 
Kahaluu 
298 
0 
0 
12.04 
Heeia 
0 
5.55 
1 
5.88 
Heeia 
0 
5.55 
0 
5.88 
Keaahala 
0 
10.61 
31 
1.08 
Keaahala 
1 
10.61 
57 
1.08 
Kaneohe 
1 
11.58 
947 
0 
Kaneohe 
0 
11.58 
392 
0 
Total 
603 
980 
Total 
313 
449 
