22 
Fishery Bulletin 120(1) 
Edge = 
0.013 x Length —- 0.075 
r=0.85 
E 
E 
S 
® 
= 
is 
Qa 
Total length (cm) 
Figure 8 
Relationships between the diameters of marks on illicia and total lengths of 
goosefish (Lophius americanus) captured along the Atlantic coast from Virginia 
to Georges Bank from May 2015 through July 2017. Multiplication signs, cir- 
cles, and crosses indicate values for diameters of settling checks, first rings, and 
edges of illicia, respectively. For each type of mark, trend lines, linear regression 
equations, and coefficients of correlation (r”) are shown. 
low precision of readings of illicia in our study. Other rele- 
vant factors may be the age reader’s limited experience 
with the structure and possible differences in lighting 
between the 2 readings. The age reader did not identify 
the same rings and may have picked a different position 
for the first ring each time. However, the reader was care- 
ful to avoid counting a false annulus, as is seen in other 
Lophius species. This false annulus was identified in 14 of 
the 22 images of illicia from goosefish annotated by 
J. Landa (Landa’). 
Recommendations for reading illicia of other Lophius 
species (Duarte et al.') include starting with larger (>50 cm 
TL) fish and adjusting the lighting and focus as each sam- 
ple is viewed. The fish in this study were mainly smaller 
than 50 cm TL, precluding the first suggestion, although 
the samples used by Bank et al. (2020) were mainly over 
50 cm TL. The age reader in our study adjusted the micro- 
scope while viewing samples; however, it was impossible to 
do so during training on images. 
The age reader’s vertebral ring counts never agreed 
with known ages and were always higher than the known 
age (often by 2-3 years). Overestimation of ages derived 
First ring = 
0.0017 x Length + 0.20 
from vertebrae also occurred for the 
2 smallest fish in the study by Bank 
et al. (2020). Possible reasons for the 
overestimation of vertebral ages include 
checks due to sporadic feeding behavior, 
variation in short-term growth rates, 
and variation in temperature as the fish 
migrate vertically (Bank et al., 2020). 
In our study, precision of vertebral ring 
counts was below acceptable levels but 
was much higher than ring counts from 
illicia. The differing levels of precision 
may have been because of the age read- 
er’s greater level of experience with ver- 
tebrae than with illicia or because of the 
false annuli and multiple checks seen in 
illicia. 
Marginal increment analysis 
Results from marginal increment anal- 
ysis (Fig. 7) provide evidence of annual 
formation of rings on the illicitum. How- 
ever, the analysis was not conclusive as a 
result of low sample sizes, high variabil- 
ity, and a short time span (13 months). 
This analysis included only fish of ages 
from 1+ to 2+ years; therefore, these 
results should not be applied to older 
age classes. 
We followed Campana’s (2001) guide- 
lines for implementing marginal incre- 
ment analysis as much as was feasible: 
we made an effort to view samples in 
random order and examined only one 
year class. Our sampling did not extend 
for 2 years, as recommended, and use of 
larger sample sizes may have revealed a cyclic pattern. 
However, our results provide a provisional estimate of the 
timing of ring formation. 
Growth 
Growth rates estimated from modal length progression 
of the 2015 year class (11-16 cm TL/year; Table 1) were 
much faster than estimates based on readings of vertebrae 
or illicia in our study (Fig. 9). Our illicium-derived growth 
estimates (7-10 cm TL/year) were slightly faster than our 
vertebra-derived estimates (5-10 cm TL/year). The verte- 
bral estimates were similar to those from earlier studies 
in which vertebrae were used (6-10 cm TL/year; Richards 
et al., 2008), indicating that the aging method was applied 
similarly in the 2 time frames and that the 2015 year class 
grew normally. 
Our estimate of growth in the first few months of life 
(1.52 mm/d) is comparable to the rates of 1.4 mm/d 
(pelagic) and 1.3 mm/d (benthic) for young-of-the-year fish 
found by Able et al. (2007), who had a larger sample size 
(n=60) than that of our study. However, our observations of 
