Nixon and Jones: Age and growth of larval and juvenile Micropogonias undulatus 
775 
an otter trawl with a 9. 1-m lined net containing 15.9- 
mm mesh and a 6.4-mm mesh liner. Otter trawls were 
towed at 1.0 to 1.5 m/s over the bottom for five min- 
utes. Specimens were preserved in 70% ethanol im- 
mediately upon collection. The ethanol was changed 
within 24 hours and again after two days. 
Otolith processing and data analysis 
Standard length (SL) measurements were made on 
fish to the nearest 0.1 mm with an image analysis 
system for individuals <20 mm SL or with vernier 
calipers for individuals >20 mm SL. Sagittal and 
lapillar otoliths were extracted from at least 30 fish 
chosen at random from each station and sampling 
date (rc=605, 40 from the MAB and 565 from estua- 
rine waters). Otoliths were extracted from all indi- 
viduals when samples contained less than 30 fish. 
Only 40 of the 126 larvae collected in the MAB were 
available for age analysis owing to inadequate pres- 
ervation. Otolith maximum diameter (OMD) was 
measured on sagittae from rostrum to postrostrum 
to the nearest 0.1 mm with an image analysis sys- 
tem — 39 otoliths from larvae collected in the MAB 
and 143 otoliths from randomly selected estuarine 
larvae and juveniles were measured. 
The right sagittal otolith was used in age and 
growth analyses except when lost or damaged; then 
the left otolith was used. Procedures for the prepa- 
ration of otoliths that required sectioning and pol- 
ishing followed Epperly et al. (1991) — in short, 
otoliths were sectioned longitudinally, ground, and 
then polished to the primordia on both sides. Gener- 
ally, otoliths from fish <15 mm SL did not require 
grinding or polishing to distinguish daily increments; 
they were placed directly on glass slides and embed- 
ded in Euparal. Otoliths were read at l,000x, under 
cross-polarized, transmitted light on a monitor with 
an image analysis system. All specimens were aged 
without knowledge of fish size or collection date. 
Three independent age counts were averaged to de- 
termine final ages. Age counts were estimated by 
adding 5 days to the number of daily increments in 
the otoliths by assuming that increment deposition 
begins at 5 days posthatching as in spot (Leiostomus 
xanthurus', Peters et al. 1 2 ). 
1 Dameron, J. C., P. J. Geer, C. F. Bonzek, and H. M. Austin. 
1994. Juvenile finfish and blue crab stock assessment pro- 
gram, bottom trawl survey. Annual data summary report se- 
ries vol. 1987. Special Scientific Report 124, Virginia Insti- 
tute of Marine Science, College of William and Mary, Gloucester 
Pt„ VA. 
2 Peters, D. S., J. C. Devane Jr., M. T. Boyd, L. C. Clements, and 
A. B. Powell. 1978. Prebminary observations of feeding, growth, 
and energy budget of larval spot ( Leiostomus xanthurus). In 
Annu. Rep. NMFS, Beaufort, NC, p. 377-397. 
Paired /-tests were used to determine if there were 
significant differences in age estimates between 
lapillar and sagittal otoliths (n=32) and in size and 
age counts between left and right otoliths (rc=30). 
Also, the precision of sagittal age counts by the pri- 
mary reader and a secondary reader were compared 
(ti= 50 ) with the indices of average percent error 
(Beamish and Fournier, 1981), coefficients of varia- 
tion, and index of precision (Chang, 1982). A paired /- 
test also was used to determine if there was a signifi- 
cant difference in mean age counts between readers. 
To generalize comparisons of mean growth rates 
and size-at-age of Atlantic croaker across capture 
sites, stations were grouped geographically into re- 
gions. These regions were designated as MAB, Chesa- 
peake Bay, seaside Eastern Shore (includes the 
Wachapreague and Sand Shoal Channel stations), 
bayside Eastern Shore (includes the Occohannock 
Channel station), marshes (includes the Tue and 
Guinea marsh stations), and rivers (includes the 
James and York river transects). The length and age 
of fish were compared among regions sampled with 
similar gears with independent, two-sample /-tests. 
Linear regression comparisons (Rawlings, 1988) 
were used to compare growth rates (slopes) and size 
at day 0 (y-intercepts) between early- (September 
through October) and late-captured (November 
through March) larvae <15 mm SL and <80 d. The 
analysis was restricted by size because larger, older 
juveniles were not available during early-season col- 
lections. Linear regression comparisons (Rawlings, 
1988) were also used 1) to compare growth between 
early- (July through August) and late-season (Sep- 
tember through February) spawned larvae (<19 mm 
SL) and 2) to compare growth between early- and 
late-season spawned juveniles (19.1-65 mm SL). 
Early- and late-spawned larvae and juveniles were 
analyzed separately so that linear growth patterns 
could be described for the two life stages. We also 
used ANCOVA to compare mean size between early- 
and late-spawned juveniles. 
A Laird-Gompertz growth model (Laird et al., 1965) 
was used to describe the growth of Atlantic croaker 
larvae and juveniles <50 mm SL and <142 d: 
SL U) = SL (0) /expj[A (0) /a][l - exp (-a/)]); 
SL {t) = standard length at day /; 
SL (0) = assumed standard length at hatching (/=0); 
A (0) = specific growth rate at hatching (Z=0); and 
a = rate of exponential decay of the specific 
growth rate. 
The model was fitted by an iterative, nonlinear least- 
squares procedure. Age-specific growth rates were 
subsequently calculated as 
