30 
Fishery Bulletin 1 14(1) 
Water temperature (°C) Salinity 
Net water transport (m 3 /s) 
Figure 7 
Partial plots of the mixed model showing growth rate data as a function of hydrodynamic parameters for larval and early 
juvenile Atlantic croaker ( Micropogonias undulatus) collected from October 2006 through March 2007 and from September 
2007 through March 2008 in Bayou Tartellan, Louisiana. Growth rate data were obtained from directly measured otoliths 
and those ages provided from age-length keys. A static value for the other model terms was assumed for all partial plots 
in the graphical evaluation. (A) Growth rate as a function of water temperature, delineated by tide. (B) Growth rate as a 
function of salinity, delineated by tide. (C) Growth rate as a function of net water transport, delineated by tide. 
and estuarine ingress. The digital image filters reduced 
the need for human interpretation and allowed direct 
measurement and averaging of multiple readings to 
avoid aliasing from subdaily increments or other small 
scale microfeatures that could confound accurate age 
determination. Furthermore, the digital analysis al- 
lowed us to obtain exact measurement of otolith ring 
widths and distances from the core to confirm the esti- 
mated estuarine ingress date determined from changes 
in growth rates. Analyses of digital images of otoliths 
from Atlantic croaker allowed a more confident estima- 
tion of the age at which larval Atlantic croaker tran- 
sition from the continental shelf to the more hydro- 
dynamically variable and potentially more productive 
coastal boundary zone and lower-salinity estuarine 
waters. 
The linear growth model was useful for comparison 
with results from previous studies but less effective 
than the Laird-Gompertz growth curves in detailing 
accurate growth rates of Atlantic croaker larvae. The 
Laird-Gompertz growth models allowed hindcasting 
to accurately include the NL at hatching and rate of 
growth more effectively in the dah before sampling. 
They provided more detail about the timing of ingress 
of Atlantic croaker into estuaries, where growth rates 
were expected to increase. The models revealed the lim- 
ited somatic growth of larvae before recruitment into 
the lower-salinity estuarine system of Bayou Tartellan. 
Moreover, they allowed us to calculate instantaneous 
growth rates that reflect small-scale, daily changes af- 
fected by spatial location within the recruitment cor- 
ridor without the bias introduced from overall aver- 
ages expressed in a linear relationship with a singular 
rate of 0.20 inrn/d or without the bias introduced from 
groupings of larvae in dah. 
With the Laird-Gompertz models we estimated maxi- 
mum growth rates, showing the difference of maximum 
growth rates that occurred later (with respect to dah) 
