Leber et a I.: Influence of release season on size-dependent survival of Mugil cephalus 
273 
and direct relationship between size-at-release and 
recapture rate (Fig. 4) — when fish were released in 
summer, recapture frequency was almost directly 
0 10 20 30 40 50 
Weeks after release 
Figure 3 
Mean total length (± SE) of cultured fish recaptured in 
collections made following spring and summer releases into 
Kaneohe Bay. Data are for the 70-85 mm size-at-release 
interval. Length was averaged within replicate release lots. 
Standard errors were based on replication established by 
release lots (n = 6 lots per season [3 at each release site], 
not total number of individuals recaptured). 
2 ° 
Figure 4 
Recapture frequencies of tagged cultured Mugil cephalus 
recaptured in cast-net samples after summer releases into 
Kaneohe Bay. Data are presented for each of the five size 
intervals released (size-at-release: 1=40-60 mm total 
length, 2=60-70 mm, 3=70-85 mm, 4=85-110 mm, and 
5=110-130 mm). Data are given as percent recaptured fish 
of the total fish released per size-at-release interval. 
proportional to size-at-release within 1 month after 
release. This pattern was evident throughout the rest 
of the study. In contrast, size-at-release had much 
less effect on recapture frequencies for fish released 
10 weeks earlier, in the spring (Fig. 5). 
Recapture frequencies of small tagged fish ( <70 
mm TL) were clearly greater throughout collections 
made following spring releases than in those after 
summer releases. After 45 weeks in the wild, fish 
from the smallest size classes released in spring re- 
mained abundant in net samples. The relative im- 
pact derived from the smallest fish released in spring 
(45-60 mm) corresponded to impacts of some of the 
larger sizes released. In contrast, on the majority of 
collection dates following summer releases, not a 
single individual (released in summer) was collected 
from the 45-mm to 60-mm size-at-release group. After 
a few months in the wild, the larger fish released ( >85 
mm) generally were more abundant in samples when 
they were liberated in summer rather than in spring. 
To compare recapture frequencies statistically 
among size-at-release intervals, values per release 
lot were summed across weeks for the period between 
16 and 36 weeks after releases. After summer re- 
leases, mean recapture frequencies of fish <70 mm 
when released were substantially less than frequen- 
cies for fish > 85 mm when released (Fig. 6; ANOVA, 
P < 0.001 in a posteriori orthogonal contrasts [Sokal 
and Rohlf, 1981] of intervals 1 and 2 combined ver- 
sus intervals 4 and 5 combined). 
