624 
Fishery Bulletin 99(4) 
was a graphical rather than a statistical exercise (Labelle 
et al., 1993; Thompson et ah, 1999), results corroborate 
otolith-based estimates of growth. 
Direct comparison of linear growth functions computed 
for tagged fish and otolith-aged fish provides further sup- 
port for otolith-based estimates of growth (slopes from 
the two equations differed by only 0.002 mm/d). Francis 
(1988) cautioned against direct comparison of age-based 
growth estimates with growth estimates from tagging 
data because different information results from the dif- 
ferent data types. The linear model fitted to length-at-age 
data predicts TL of fish for a given age (in d), whereas the 
linear model fitted to tagging data predicts the increment 
of growth expected of a fish at liberty for a given number of 
days. However, because the relationship between TL and 
age is linear over the range of TL from the tagging data 
(Szedlmayer and Shipp, 1994), we submit that the slopes 
can be compared as estimates of mean growth. The slope 
of the age-based linear model is an estimate of the pop- 
ulation growth rate of young fish, whereas the slope of 
the length-based linear model is an estimate of the mean 
growth of tagged individuals. That growth estimates for 
otolith-aged and tagged red snapper are very similar cor- 
roborates the aging method with otoliths and indicates 
that on average tagging did not affect fish growth. 
Red snapper off Alabama 
The red snapper fishery off Alabama is unique in several 
ways. Despite the reported lack of high-relief natural reef 
structures on the continental shelf off Alabama (Parker et 
al., 1983; Shultz et al., 1987; Schroeder et al., 1989), fish- 
ermen land a disproportionately high percentage of the 
annual GOM red snapper harvest from Alabama coastal 
waters (Schirripa and Legault, 1999). Although Alabama 
red snapper landings have accounted for approximately 
11% of total GOM landings from 1981 to 1998, the recre- 
ational fishery off Alabama accounted for mean (±SE) 21.1 
(±2.25)% of GOM recreational landings from 1981 to 1998 
and 26.5 (±1.83)% of recreational landings from 1990 to 
1998 (Schirripa and Legault, 1999). Total landings of red 
snapper from the U.S. GOM averaged 3.5 x 10 3 metric tons 
from 1981 to 1998, with commercial and recreational land- 
ings essentially equal; however, 94% of Alabama red snap- 
per landings from 1990 to 1998 came from the recreational 
sector of the fishery (Schirripa and Legault, 1999). 
The red snapper fishery off Alabama is also unique be- 
cause it is prosecuted almost entirely over artificial reefs 
(Minton and Heath, 1998; Shipp, 1999). Artificial reef con- 
struction began in the 1950s when charter boat operators 
gained permission from the Marine Resources Division of 
the Alabama Department of Conservation to place 250 car 
bodies on the sea floor off Alabama (Hosking and Swingle, 
1989; Minton and Health, 1998). Since the 1950s, artifi- 
cial reefs have been constructed of a variety of materials 
including car bodies, liberty ships, army tanks, kitchen 
appliances, newspaper bins, and most recently, prefabri- 
cated concrete structures. Reef building increased dra- 
matically in the 1980s with the creation (by permit) of ar- 
tificial reef areas that now encompass a total of 3100 km 2 , 
