Frazier et al.: Growth rates of Sphyrna tiburo estimated from tag-recapture data 
Table 3 
Estimates of von Bertalanffy growth function (VBGF) parameters, maximum age, and age at 50% maturity from GROTAG models 
based on tag-recapture data and from age-based models for male and female bonnetheads (Sphyrna tiburo) from the northeast- 
ern Gulf of Mexico (GOM) and the Atlantic Ocean off the southeastern United States (Atlantic region). The VBGF parameters, 
asymptotic length (L..), coefficient of growth (k), and theoretical age at length zero (¢)), are provided with 95% confidence intervals, 
which were generated by using bootstrapping (with 5000 iterations). Also provided are published estimates of maximum age 
(Lombardi-Carlson et al., 2007; Frazier et al., 2014). Ages at 50% maturity were estimated in GROTAG models by using lengths at 
50% maturity published in Lombardi-Carlson et al. (2007) and Frazier et al. (2014). Data used in the models are from sharks tagged 
and recaptured in the GOM during 1993-2006 and in the Atlantic region during 1998-2019. FL=fork length. 
Model 
GROTAG 
Atlantic region 
Age based 
Atlantic region 
L., 
(mm FL) 
769.6 
604.2—823.5 
948.3 
921.2—967.9 
814.8 
533.4-1135.6 
1039.5 
1027.7—1048.5 
703.3 
663.8—765.2 
895.0 
842.9-970.3 
7179.8 
761.5—799.7 
1032.3 
1011.6—1053.9 
k 
0.254 
0.470-0.087 
0.243 
0.313-0.184 
0.166 
0.386—0.070 
0.170 
0.191-0.145 
0.538 
0.753-0.386 
0.282 
0.334—0.226 
0.296 
0.320-0.274 
0.188 
0.179-0.198 
to 
(years) 
-1.56 
—(1.14—4.19) 
-1.27 
—(1.02-1.64) 
—2.35 
—(1.76-3.76) 
-1.75 
—(1.54-1.99) 
-1.60 
—(1.10—2.25) 
-2.13 
—(1.82—2.55) 
-1.51 
—(1.35—-1.68) 
-1.76 
—(1.59-1.94) 
Age at 50% 
maturity 
(years) 
5.2 
oo12.5 
4.5 
3.85.7 
6.2 
oo—7.5 
7.4 
6.8-8.5 
2.0+ 
3.0+ 
3.9 
6.7 
Maximum 
age 
(years) 
8.9 
oo—15.3 
11.4 
9.7-14.3 
21.5 
19.4—co 
24.0 
23.5-25.8 
6.5 
6.4-6.6 
10.8 
10.2-12.1 
18.7 
16.8—co 
23.0 
21.9-24.9 
Published 
maximum 
age 
(years) 
occurs (Ulrich et al., 2007). In the GOM, the lower num- 
ber of male bonnetheads that were recaptured is a result 
of the primary habitat of male bonnetheads being outside 
of or on the periphery of the area where sampling occurred 
and where tags were deployed. Therefore, fewer males than 
females were tagged and consequently recaptured. Future 
efforts should focus on sampling and tagging male bonneth- 
eads, especially those in early life stages, to decrease uncer- 
tainty in results from length-based models. 
Although it is difficult to determine the effects of tagging 
on growth in wild fish populations, our results do not pro- 
vide evidence indicating that tagging and tag type affected 
growth. Growth was variable in the first year after tagging, 
with a larger number of individuals experiencing slower 
growth than expected. This slow growth, however, could be 
due to the short-term effects of the stress from being cap- 
tured rather than the effect of the tag on the individual 
(Gruber, 1982; Parsons, 1987; Davenport and Stevens, 1988; 
Skomal and Bernal, 2010). Over the long term, we found no 
evidence of effects of tagging on growth of captured bonnet- 
heads, and all individuals at liberty for over 4 years grew 
older and larger than the age and length predicted by the 
age-based growth model. These findings contrast with those 
of studies on lemon sharks (Negaprion brevirostris) (Manire 
and Gruber, 1991; Oliveira, 2001) and northern pike (Esox 
lucius) (Scheirer and Coble, 1991) that indicate that growth 
could be impaired by up to 50% in tagged individuals when 
a variety of tag types were used. Results of other growth 
studies (e.g., Jensen, 1967; Jolivet et al., 2009) indicate 
that there were no significant effects of tagging on growth; 
therefore, it is possible that there may be effects specific 
to species, life stage, or tag type. Caution should be used 
in interpreting growth data from recaptured individuals to 
the population level without investigating these effects. 
The final GROTAG models for males in both regions 
included SD of measurement error and growth variability, 
and the GOM model has the added parameter of mean 
measurement error. Estimates of growth variability are 
unreliable because of low sample size, and 95% CIs con- 
tain values that are not above and below parameter limits; 
therefore, no comparisons can be made with these param- 
eters. The final GROTAG models for males and females in 
the GOM both have large estimates of SD of measurement 
error compared with those from the models for males and 
females in the Atlantic region; however, this result was not 
unexpected. For individuals from the GOM, length was 
measured to the nearest half centimeter, and several biol- 
ogists and interns measured fish; whereas, bonnetheads 
from the Atlantic region were measured to the nearest 
millimeter by only 2 individuals. The use of fewer individ- 
uals to measure all captured specimens and the smaller 
measurement increment likely led to more precise 
