26 
Fishery Bulletin 1 14(1) 
Age (dah) 
Figure 4 
(A) Boxplots from a comparison of estimates of standard lengths 
(SL) and ages in days after hatching (dah), by season, for otoliths 
of larval Atlantic croaker ( Micropogonias undulatus ) collected from 
October 2006 through March 2007 (year 1) and from September 
2007 through March 2008 (year 2). Fall sampling was conducted 
during September-December and spring sampling occurred during 
January-March. Nonoverlapping notches between any 2 boxplots 
represent “strong evidence” of statistically different median val- 
ues, as described by Chambers et al. 1983. (B) Relationships of SL 
versus age determined with the Laird-Gompertz growth models for 
otoliths of larval Atlantic croaker, by season and sampling year. All 
models are forced through the intercept at the estimated 1.5-mm 
notochord length at hatching to accurately reflect growth rates at 
ages in dah less than the minimum otolith determined age. Model 
parameterizations were calculated as follows: 
„2.89159(l-e _002156Age ) . 
Fall of year 1: SL = 1.5e z 
Fall of year 2: SL = 1.5-e 
Spring of year 1:SL = 1.5-e 
2.50594(1— e _002893Age ) . 
2.54838(1— c~°' 02152Age ) . 
Spring of year 2: SL = 1.5 e 
2.112601 1— <r°' 04726 -*8®) . 
rates in the fall occurred when larvae were 
approximately 20 dah younger than larvae 
whose maximum growth rates occurred in 
spring. The maximum instantaneous growth 
rate was higher in fall 2006 (0.214 mm/d; 49 
dah) than in fall 2007 (0.177 mm/d; 37 dah), 
but larvae in the fall of each year had the 
same minimum growth rate of 0.093 mm/d 
(Fig. 5). A greater initial instantaneous 
growth rate was observed in spring 2008 
(0.150 mm/d) than in spring 2007 (0.113 
mm/d). Although, the maximum instanta- 
neous growth rate in spring 2008 (0.216 
mm/d) was greater than the maximum rate 
in spring 2007 (0.196 mm/d), it occurred at 
an earlier age (16 dah versus 32 dah) and 
decreased rapidly thereafter. The 10-d aver- 
ages for the 2 sampling years were slightly 
different. The highest average growth rate 
in year 1 was 0.203 mm/d at ages between 
40 and 50 dah (Table 2), correlating with 
a peak at 49 dah in the fall of sampling 
year 1 (Fig. 5). In year 2, the highest aver- 
age growth rate of 0.193 mm/d occurred at 
ages between 20 and 30 dah (Table 2) — a 
rate that was consistent with the peak in 
instantaneous growth rates that occurred at 
16 dah during the spring of year 2 (Fig. 5). 
The estimated average date of estuarine 
ingress after hatching was similar for both 
fall and spring seasons in both sampling 
years, but there were differences between 
seasons in the mean distance of the ring 
from the otolith core. Regardless of season 
or year, rapid changes in distance of the 
ring from the core and ring width happened 
around 40 dah, before the ontogenetic shift 
from late larvae to early juvenile has been 
reported to occur at approximately 10 mm 
SL (Barbieri et al., 1994a, 1994b). Mean 
distance of the otolith ring from the core, 
as a proxy for growth, was fairly stable be- 
fore 40 dah and increased more rapidly and 
became more variable after that point (Fig. 
6A). In particular, in spring 2007 and 2008, 
distances from the core were greater than 
the distances for the same age in the fall of 
2006 and 2007. Mean otolith ring width was 
somewhat constant at approximately 0.5 pm 
for the fall of both sampling years and 1.1 
pm for the spring of both years until 40 
dah, when mean ring width became much 
more variable and generally increased with 
ring count thereafter (Fig. 6B). 
Statistical analysis 
The mixed model fitted for water tempera- 
ture, salinity, and NWT, by tide, to the es- 
timated growth rate of all individuals col- 
