Kupchik and Shaw: Age, growth, and recruitment of larval and early juvenile Micropogonias undulatus 
27 
Fall — Year I 
Fall— Year 2 
Spring — Year 1 
Spring — Year 2 
0.25 n 
0.20 - 
0.15 
0.10 
0.05 
Age (dah) 
Figure 5 
Maximum growth rates for Atlantic croaker ( Micropogonias undu- 
latus) larvae and early juveniles collected in Bayou Tartellan, Loui- 
siana, from October 2006 through March 2007 and from September 
2007 through March 2008, based on the first derivative of the Laird- 
Gompertz growth model for otoliths of larval Atlantic croaker. Solid 
lines indicate daily growth rates (mm/d I as the slopes of the estimates 
from the Laird-Gompertz models at any age in days after hatching 
(dah). Dashed lines indicate the maximum growth rates for fall of year 
1 (2006), fall of year 2 (2007), spring of year 1 (2007), and spring of 
year 2 (2008). 
lected during both sampling years, having significant 
model terms for all continuous variables. The only 
model terms that were not significant at the a priori 
alpha level of 0.05 were tidal stage and salinity as a 
function of tidal stage. The nonsignificance of the tidal 
stage term is likely a function of masking due to higher 
order interaction terms. 
The individual model terms generally showed a 
trend of increased growth rate with ingress through 
the tidal pass, supporting the notion that otolith ring 
width and growth rate both increase at ingress, regard- 
less of sampling year or season. The interaction of ebb 
tides with water temperature showed higher growth 
associated with colder temperatures that are indicative 
of shallow estuarine waters during late fall and win- 
ter (October-March) (P=0.040; Fig. 7A). The flood tide 
interaction with water temperature showed a similar 
trend, although more muted, likely a result of high- 
er temperatures of waters from the inner continental 
shelf to the coast. Increased growth was also associ- 
ated with decreasing salinity, regardless of tidal stage, 
resulting in a nonsignificant interaction (P=0.901), al- 
though the 95% confidence interval was 
much larger at lower salinities during 
flood tides (Fig. 7B). As with interactions 
with water temperature, modeled growth 
rates associated with salinities that were 
more consistent with estuarine condi- 
tions were significantly greater (P=0.046) 
than growth rates associated with higher 
salinities. Higher growth rates also were 
associated with large negative NWT on 
ebb tides, indicative of lower-salinity wa- 
ters further up the estuary (P=0.001; Fig. 
1C). However, during flood tides, no rela- 
tionship was apparent statistically. 
Discussion 
Successful estuarine recruitment of At- 
lantic croaker larvae through tidal pass- 
es along the northern GOM depends on 
a highly variable spawning regime, on 
advantageous environmental conditions 
such as hydrographic, tidal, and wind- 
forcing factors, and ultimately on lar- 
val growth (Norcross and Austin, 1988; 
Eby et al., 2005; Montane and Austin, 
2005). Hindcasting of the Laird-Gom- 
pertz growth model to 0 dah allowed us 
to infer growth rates of larvae along the 
recruitment corridor across the continen- 
tal shelf and coastal zone. The bottleneck 
nature and the highly variable hydrody- 
namic environment of a seasonally well- 
mixed tidal pass at Bayou Tartellan, as 
well as the associated increase of larval 
growth with ingress into the estuarine 
nursery ground (Searcy et al., 2007), pre- 
sented the challenges and rewards of successful estua- 
rine recruitment for larval Atlantic croaker that were 
spawned offshore. 
The highest frequency of hatching dates occurred 
between late September and early October for both 
sampling years, but hatching continued through the 
late winter and early spring (Fig. 5), indicating an 
overwinter spawning and recruitment period. The peak 
hatching dates corresponded well with the previously 
described period of July through December for peak 
spawning and recruitment (Warlen and Burke, 1990; 
Barbieri et al., 1994b) and with an overall spawning 
and recruitment period from August through May (Het- 
tler and Chester, 1990). Differences in the distribution 
of hatching dates between year 1 and year 2 of the 
study highlight the variability in yearly spawning of 
Atlantic croaker that was due to factors on various 
spatial (Miller and Able, 2002) and temporal (Norcross, 
1983) scales. For example, year 2 distribution of hatch- 
ing dates peaked less than that of year 1. In addition, 
a higher percentage of larvae were recruited in the 
months after December 2007 than in other periods, 
