Fishery Bulletin 120(1) 
vertebra shows 4 rings. 
Rings on each structure were counted twice following 
NEFSC aging protocols (Penttila and Dery, 1988) by one 
age reader, with at least 2 weeks between readings. Data on 
month, but not year, of capture were available to the age 
reader. Any samples for which the 2 ring counts differed 
were reexamined later; samples for which the count still 
could not be resolved were excluded from the analysis. 
The reading order of rows of illicia (10-15 samples per 
row) was randomized before samples were examined. Verte- 
bral samples were viewed in a haphazard order, by mixing 
together the envelopes that held the samples and viewing 
the samples in the order they were pulled out of a box. 
The reader had over 15 years of experience with age 
reading methods, including the vertebral method for aging 
goosefish. However, the reader had examined fewer than 
200 illicia prior to this study. Immediately before exam- 
ining the study samples, the reader reviewed a set of 24 
images of illicia from goosefish that had been examined 
and annotated by another age reader (Landa’) experi- 
enced with illicia from anglerfish sampled in Europe. 
Sectioned illicia (Fig. 3A) were immersed in a 1:1 solution 
of glycerin and 70% isopropyl alcohol and were viewed with 
transmitted light at 40x magnification on an Olympus BH2 
compound microscope (Olympus Corp.). Recommended 
practices for viewing illicia of white anglerfish (Duarte 
et al.'; Landa et al., 2013) were followed, including adjust- 
ing lighting and focus, counting the dark rings, and using 
axes with greater color contrast. It was assumed that the 
° Landa, J. 2016. Personal commun. Cent. Oceanogr. Santander, 
Inst. Esp. Oceanogr., Promontorio San Martin s/n, 39004 
Santander, Cantabria, Spain. 
Figure 3 
Annotated images of age structures, (A) an illicium section and (B) a baked vertebra, taken from a 41-cm-TL 
goosefish (Lophius americanus) with a known age of 2 years that was captured on 9 April 2017 south of Rhode 
Island. Each ring is marked with a red dot. The image of the illicium section shows 2 rings and the following 
measured diameters: settling check (red), first ring (green), last ring (blue), and total (pink). The image of the 
first ring would be nearly circular and that it would be out- 
side both an oval settling check and a false annulus (as has 
been observed in white anglerfish by Wright et al., 2002). 
This process yielded first rings comparable in diameter to 
those in the annotated images. 
Vertebrae (Fig. 3B) were interpreted following the 
methods of Armstrong et al. (1992). Samples were viewed 
with reflected light at 10x magnification on a dissect- 
ing microscope (Leica MZ6, Leica Camera AG, Wetzlar, 
Germany). 
After both structures from all fish had been assigned 
a final ring count, accuracy of readings (ring counts ver- 
sus known ages) for each structure was calculated across 
all fish. Known ages were based on the calendar year of 
capture (i.e., a fish captured in 2016 had a known age of 
1 year). Later, test counts were made on a random subset 
of samples for each structure, allowing intra-reader preci- 
sion (test counts versus ring counts) to be estimated. 
Statistics calculated for each of these measures were 
percent agreement and the mean coefficient of variation 
(CV) (Chang, 1982). For both precision tests, a Bowker’s 
test of symmetry (Bowker, 1948; Hoenig et al., 1995) was 
additionally applied to test for bias in the paired ring 
counts (one count each from the initial and the test count 
for a given structure of a sampled fish). At the NEFSC 
aging laboratory, 80% is considered an acceptable level of 
agreement; an acceptable CV is under 5% (NEFSC°). 
6 NEFSC (Northeast Fisheries Science Center). n.d. Quality assur- 
ance and quality control estimates for the production ageing of 
northwest Atlantic species: statistical measures. [Available from 
website, accessed November 2021.] 
