158 
Fishery Bulletin 106(2) 
Estimated contribution {%) of the Atlantic stock 
Figure 5 
Cumulative probability distributions of bootstrapped estimates of the contribu- 
tion of Atlantic stock female king mackerel (Scomberomorus cavalla) to landings 
at three south Florida sampling zones. Drop-lines indicate the 5 th , 50 th , and 95 th 
percentiles of bootstrap distributions. 
the next (Campana and Casselman, 1993). For example, 
cohort-specific discriminant function models computed 
for coral trout sampled on the Great Barrier Reef did 
a poor job distinguishing fish from another cohort to 
sampling region (34.3% to 39.7% classification success), 
a percentage that Bergenius et al. (2006) attributed to 
differences in growth rates ultimately caused by vari- 
ability in oceanographic conditions. 
Maximum likelihood estimates indicated that some 
percentage of winter landings in all three zones orig- 
inated from the Atlantic stock in both study years. 
However, bootstrapped confidence intervals indicated 
considerable imprecision around point estimates. Cu- 
mulative probability distributions of bootstraps (rc = 500) 
were broad for females and males in both study years. 
However, even at the lower end of the confidence inter- 
vals, Atlantic fish were estimated to have contributed 
greater than 20% of landings in all three zones, except 
for females sampled in zone 1 during winter 2002-03. 
Results potentially indicate that a distribution gradi- 
ent may exist; more Atlantic king mackerel may contrib- 
ute to landings from the Atlantic side (zone 3) and fewer 
Atlantic stock king mackerel contribute toward the 
GOM (zone 1). Mixing estimates for zone 2 are some- 
where in the middle, with the exception of zone 2 males 
in 2001-02. However, the sample size of king mackerel 
in zone 2 in 2001-02 generally was low, particularly for 
males, and this shortage could account for the higher 
estimate for the Atlantic stock contribution. 
Atlantic male and female king mackerel appear to 
have had similar contributions across all three south 
Florida sampling zones in winter 2001-02, but this 
