458 
Fishery Bulletin 95(3), 1 997 
Length and age at sexual maturity 
Females were considered sexually mature if vitel- 
logenic oocytes were present or if the histological 
sections appeared disorganized, highly vascularized, 
and contained widespread evidence of atresia. Docu- 
mentation of atresia followed the classification of 
Hunter and Macewicz (1985). Immature females had 
small (gonadosomatic index <0.35), well-organized 
gonads that contained little evidence of atresia. We 
interpreted the widespread occurrence of PAS-posi- 
tive melanomacrophage centers in inactive ovaries 
as evidence of past gonadal development, and we 
considered bonefish that had regressed (no vitel- 
logenic oocytes present) ovaries containing many of 
these structures to be sexually mature (Fig. IB). 
Males were considered sexually mature if the testes 
contained evidence of ongoing spermatogenesis, re- 
sidual sperm, or widespread PAS-positive melanoma- 
crophage centers associated with gonadal recrudes- 
cence (Fig. 1C). 
Sometimes distinguishing between the gonads of 
sexually immature bonefish and the regressed go- 
nads of mature fish was difficult. We reduced the 
probability of misclassifying regressed and immature 
fish by eliminating all bonefish collected during 
June-October from our analyses of age and length 
at sexual maturity. June-October was the season of 
minimal gonad development in bonefish, and most 
of the regressed bonefish in our sample were cap- 
tured during this period. By excluding the postre- 
productive months from our analysis, we eliminated 
84% of the regressed females and 63% of the re- 
gressed males in our sample. Hunter et al. (1992) 
recommended that only fish collected early in the 
spawning season be used to estimate the length at 
50% maturity, but our sample size was not large 
enough to allow us to restrict our analysis to this 
extent. 
To describe age and length at sexual maturity, we 
used nonlinear regression procedures to determine 
the inflection point of a logistic function fitted to the 
percentage of males and females that were sexually 
mature and to their respective lengths and ages. 
Parameter b in Table 1 is the inflection point and is 
the estimate of length or age at 50% maturity. Like- 
lihood-ratio tests were used to compare the overall 
regression models and parameter estimates for males 
and females (Kimura, 1980). 
Seasonality of gonad development 
Monthly median gonadosomatic indices (GSI) of sexu- 
ally mature males and females were plotted to show 
seasonal reproductive patterns. GSI’s were calculated 
for 449 bonefish ranging from 228 to 702 mm as 
GSI = ( GWKTW - GW)) x 100, 
where GW = total gonad weight (g); and 
TW = total fish weight (g). 
We also plotted the monthly frequency of occurrence 
of the various oocytes stages that we counted for fish 
Table 1 
Percentage mature-age, percentage mature fork length, and weight-fecundity regressions for bonefish, Albula vulpes, from the 
waters of the Florida Keys. Wt = weight (g), FL = fork length (mm), AGE = age in years, FEC = fecundity. Values in parentheses 
are standard errors. 
a 
b 
Range of X 
Y 
X 
n 
(1 SE) 
(1 SE) 
r 2 
for regressions 
y = (l/(l + e ( - 
°'*- 6)) ))xl00 
% Mature 
FL 
150 
0.028 
487.6 
0.632 
228-702 
(Females) 
(0.0064) 
(8.14) 
% Mature 
AGE 
143 
1.122 
4.24 
0.445 
1-19 
(Females) 
(0.2345) 
(0.192) 
% Mature 
FL 
116 
0.545 
417.5 
0.735 
322-687 
(Males) 
(2.8227) 
(12.59) 
% Mature 
AGE 
109 
1.618 
3.60 
0.464 
2-19 
(Males) 
(0.3674) 
(0.156) 
Y = a 
+ bX 
log 10 FEC 
log 10 Wt 
33 
1.936 
1.131 
0.706 
1,790-5,790 
(0.4708) 
(0.1312) 
