776 
Fishery Bulletin 95(4), 1997 
A( t ) = -A(o) exp (-at). 
Finally, ANCOVA was used to compare mean 
otolith size between early-captured, fast-growing 
Atlantic croaker with late-captured, slower-growing 
Atlantic croaker between 11 and 37 mm SL. 
Results 
Otolith analysis 
Age counts in sagittal and lapillar otoliths were sig- 
nificantly different (/-test, P=0.002). For older age 
fish, lapillar counts underestimated sagittal counts, 
and the disparity increased with increasing age (Fig. 
2). Also, no differences were found among size (Gtest, 
P=0.49) and age counts (£-test, P= 0.85) between left 
and right sagittal otoliths (n=30). 
Sagittal age counts by the primary reader showed 
good precision — the average percent error (APE) of 
counts was 4.8%, with a coefficients of variation (CV) 
and index of precision (D) of 6.4% and 3.8%, respec- 
tively. For the second reader’s age counts these indi- 
ces were 8.4% (APE), 11.5% (CV), and 6.7% (D). Al- 
though the second reader’s age counts had relatively 
low precision, there was no significant difference in 
mean age counts between readers (Gtest, P=0.27). 
Size and age distributions 
Monthly length- and age-frequency distributions 
showed that size and age of fish generally increased 
from September to January (Fig. 3). Size and age 
appeared to decline in February (although sample 
sizes were small) and mostly represented fish col- 
lected in the rivers after January. Also, length dis- 
tributions were highly variable in comparison with 
respective age distributions (Fig. 3). For example, 
fish collected in November had two distinct length 
modes, whereas their age frequencies clearly had only 
one mode. This pattern was also evident for fish col- 
lected in October and January; thus size does not 
appear to be a good predictor of age in these fish. 
Generally, smaller and younger fish were found in 
the seaside Eastern Shore region compared with the bay- 
side Eastern Shore or marsh regions (Mests, P<0.001) 
over the entire sampling season. This pattern was evi- 
dent regardless of gear type. Significantly smaller 
(P=0.02) and younger (P<0.001) fish were found in the 
mainstem Chesapeake Bay compared with the rivers 
inland of the Bay. However, the rivers were sampled 
only during the later half of the sampling period. 
The age of larval Atlantic croaker entering Virginia 
nursery grounds was examined from specimens col- 
lected at the mouth of the Chesapeake Bay (the most 
seaward station along the Chesapeake Bay transect) 
and at Wachapreague and Sand Shoal Channels. The 
youngest larvae (24 d) entered the mouth of the 
Chesapeake Bay on 21 September 1987 and mea- 
sured 6.1 to 7.6 mm SL (n= 3). Fish collected at 
Wachapreague and Sand Shoal Channels were prob- 
ably better representatives of the age of larvae that 
enter Virginia nursery grounds because smaller mesh 
nets (with 3.2-mm mesh liner which sampled smaller 
larvae more effectively) were used at these stations. 
The youngest larvae observed at Wachapreague 
Channel were collected on 29 September 1987 with 
a mean size and age of 7.3 mm SL and 26 d, respec- 
tively, with the youngest individuals (20 d) measur- 
ing 5.4 and 6.1 mm SL (n= 2). The youngest larvae 
observed at Sand Shoal Channel were collected on 
30 September 1987 with a mean size and age of 8.3 
mm SL and 29 d, respectively, with the youngest in- 
dividuals (23 d) measuring 6.1 and 7.3 mm SL (n= 2). 
In conclusion, it appeared that the youngest larvae 
entered Virginian estuaries at an approximate age 
of 20 to 26 d, measuring 5.4 to 7.6 mm SL. 
Hatching-date distributions 
Hatching-date distributions indicated a protracted 
spawning period of 8 months from 5 July 1987 to 10 
February 1988 and with 82% of spawning limited to 
