survey conducted in 2015 (average date: 7 June 2015) 
and was visible in survey catches through the autumn of 
2018. Between September 2015 and April 2018, 204 fish 
ft 7 Jun 2015 
| ] i] i] i] i] ] [] t 
20 Sep 2015 
1 May 2016 
[] [] 
29 May 2016 
1 1 
27 Sep 2016 
1 
2 Apr 2017 
1 1 
11 Jun 2017 
Proportion at length 
SPSS SS SHELLS SELELEE SEELEE SHEL SLEEEG SOEESS SOESESEE 9SO9S09 SSES99 9999099 
DO+t-]NN OO+=NN COO$]—]NDN COO$]+NN OO}]$NN ODO] =]NN ODO] $ WN OO] | NN OO} $ WN OO + | NN OO——= nono 
9 M10 W110 D1 OMNOVOUNOVN1OVIOVN OVIOV1IO V1 OVUIOVUIO V1 OVUNIOVIOVN OUOVUIO MN OUIOVUIOM OUOVIOWN OMOM0U1IOMOWIOUW 
1 1 
25 Oct 2017 
1 
7 Apr 2018 
1 Jun 2018 
22 Sep 2018 
Total length (cm) 
Figure 4 
Length—frequency distributions of goosefish (Lophius americanus) from the 2015 
year class captured along the Atlantic coast of the United States from Virginia 
to Georges Bank between June 2015 and September 2018 during scallop dredge 
(white bars) and autumn (gray bars) and spring (black bars) bottom trawl surveys 
conducted by the NOAA Northeast Fisheries Science Center. The graphs illus- 
trate the length progression of the 2015 year class. Dates indicate the average 
date of each survey, and curved lines represent the normal curves that were fit to 
stratified mean number per tow at length for the year class during each survey. 
Fishery Bulletin 120(1) 
were collected from this length mode as 
it grew (Table 2, Fig. 1). 
Figure 4 shows a new year class 
entering the population in 2018. This 
2018 year class had a modal length of 
approximately 14 cm TL in early June 
and a length of 20 cm TL by September. 
These lengths are similar to the lengths 
of the 2015 year class at age 0 during 
the same times of year (11 and 22 cm 
TL, respectively; Table 1). 
Hatch date 
Individual counts of daily rings averaged 
91.7 and 99.3 for the 2 fish caught on 4 
September 2015. There was less than a 
5% difference between the minimum and 
maximum counts from each fish (ranges: 
90-94 rings and 96-101 rings). The esti- 
mated hatch dates were 28 May 2015 
and 4 June 2015. The average growth 
rate (from the estimated hatch date until 
the date of collection) was 1.52 mm/d for 
both fish. 
Age estimation 
Ring counts were assigned to 190 illicia 
and 203 vertebrae (Suppl. Table) (online 
only). Ring counts could not be assigned 
to 14 illicia because of missing or broken 
structures (5 fish), processing problems 
(8 fish), or failure to reach a consensus 
(1 fish). For readings of vertebrae, no 
fish were omitted because of processing 
or aging issues (the spinal column was 
not collected for one of the fish sampled 
in September 2015). 
Accuracy of ring counts from illicia 
was 50.0%, with a CV of 20.0% (num- 
ber of samples [n]=190) (Fig. 5A). Where 
disagreements occurred, the ring count 
was always higher than the known age. 
In contrast, accuracy of ring counts from 
vertebrae had 0% agreement with the 
known age and a CV of 69.5% (n=203) 
(Fig. 5B). All vertebral ring counts were 
higher than known ages (98% exceeded 
known age by 2-3 years), with increasing 
error for higher known ages. 
Precision of ring counts from both 
structures was low. Agreement between 
repeated readings of illicia was only 
8.8% (CV=44.6%, n=57) (Fig. 6A), with 
a strong bias toward higher counts 
in the test set (Bowker’s test: P<0.001). Agreement 
between readings from vertebrae was 67.9% (CV=5.9%, 
n=56) (Fig. 6B). 
