212 
Fishery Bulletin 115(2) 
Table 1 
Spawning intervals for larval Gulf menhaden {Brevoortia patronus) collected 
in Bayou Tartellan, Louisiana, during 2006-2008. Spawning dates were deter¬ 
mined from back-calculated otolith ages and collection dates after application 
of age-length keys. Percentages of the total number of larvae collected in each 
sampling year and cumulative percentages for each sampling year are based 
on half-month intervals. 
Interval 
2006-2007 
2007-2008 
% 
Cumulative 
% 
% 
Cumulative 
% 
09/01-09/15 
0.07 
0.07 
0.20 
0.20 
09/16-09/30 
1.39 
1.46 
0.40 
0.61 
10/01-10/15 
15.28 
16.74 
1.97 
2.58 
10/16-10/31 
8.45 
25.19 
9.91 
12.49 
11/01-11/15 
2.95 
28.14 
10.12 
22.61 
11/16-11/30 
2.49 
30.63 
4.86 
27.47 
12/01-12/15 
6.79 
37.42 
11.33 
38.80 
12/16-12/31 
9.70 
47.12 
8.50 
47.29 
01/01-01/15 
2.49 
49.62 
5.92 
53.21 
01/16-01/31 
6.58 
56.20 
14.67 
67.88 
02/01-02/15 
20.96 
77.17 
19.47 
87.35 
02/16-02/28 
19.13 
96.29 
7.28 
94.64 
03/01-03/15 
2.25 
98.54 
3.95 
98.58 
03/16-03/31 
1.46 
100.00 
1.42 
100.00 
initial portion of this growth stanza had a maximum 
growth rate of 0.21 mm/day, and an average growth 
rate of only 0.11 mm/day. The 2-cycle Laird-Gompertz 
models were also fitted separately to year 1 (October 
2006 to March 2007) and year 2 (September 2007 to 
March 2008); however, these 2 models did not result 
in better performance over the pooled model (F-test: 
P=0.201; Fig. 3). 
Individual Laird-Gompertz models were fitted for 
pre- and postmetamorphosis, first where the stages are 
delineated by age, and secondly where the stages are 
determined from SL. The breakdown by age between 
the 2 Laird-Gompertz models provided better fits in 
describing somatic growth during each period in con¬ 
trast with the grouping determined by SL because of 
the increased variability in length at a particular age. 
The first Laird-Gompertz model for larval stage based 
on age groupings had an initial specific growth rate 
of 0.072/day, and a weaker decay rate of 0.019. This 
larval stage had a modeled maximum growth rate of 
0.71 mm/day, and a mean growth rate of 0.47 mm/day 
(Fig. 4A). During the initial portion of the transition to 
the juvenile developmental stage, the decay rate for the 
age-grouped model was 0.051. This stage in the age- 
grouped model had a maximum growth rate of 0.20 
mm/day, and an average growth rate of 0.08 mm/day. 
The premetamorphic larval stage for the SL grouping 
had an initial specific growth rate of 0.086/day, and a 
decay rate of 0.041. This premetamorphic stage had a 
modeled maximum growth rate of 0.41 mm/day, and 
a mean growth rate of 0.38 mm/day (Fig. 4B). At the 
onset of metamorphosis to the juvenile developmental 
stage, the decay rate was 0.015. The very beginning 
of the ontogenetic shift to the juvenile developmental 
stage had a comparatively lower maximum growth rate 
of 0.12 mm/day, and an average growth rate of 0.10 
mm/day. 
Analyses of otolith microstructure showed changes 
in both mean ring distance from the otolith core, as 
well as in mean ring width after 33 das. Mean ring 
distance from the core during the initial larval stage 
was similar for both October 2006 to March 2007 and 
September 2007 to March 2008. After the beginning of 
the model-estimated ontogenetic shift at approximately 
33 das, otolith growth slowed and showed limited vari¬ 
ability, and year 2 otolith growth was slower than that 
of year 1 (Fig. 5A). Mean ring width showed a similar 
pattern: a decline in ring width occurred after 33 das 
and limited differences between either sampling years. 
Before the beginning of the ontogenetic shift onset, 
ring width for both sample years appeared to increase 
slightly, agreeing with the 2-cycle modeled growth rate 
that was largest just before the expected onset of shift 
in feeding strategy (Fig. 5B). 
Discussion 
Hatching length as calculated by the 2-cycle, Laird- 
Gompertz model was 3.34 mm, which agreed well 
