170 
Fishery Bulletin 117(3) 
its diet (Navarro-Gonzalez et al., 2012) and morphology 
(Gonzalez-Isais and Montes-Dommguez, 2016; Montes- 
Dominguez and Gonzalez-Isais, 2017; Navarro-Gonzalez 
et al., 2018). Consequently, it is categorized as a data 
deficient species in the International Union for 
Conservation of Nature IUCN Red List (Valenti and 
Robertson, 2009). 
The high catch volumes of Panamic stingrays in the 
shrimp trawl fishery in the central zone of the Pacific 
coast of Colombia, the significant increase in the domi¬ 
nance of this species in the elasmobranch assemblage of 
the study area, the decrease in the catch rate (Navia and 
Mejfa-Falla, 2016), and the lack of information on its life 
history traits highlight the importance of contributing rel¬ 
evant information for future population assessments. The 
aim of this study was to estimate and compare the age and 
growth parameters of male and female Panamic stingrays, 
by using a multi-model approach and inference. 
Materia! and methods 
Specimen collection 
Individuals of the Panamic stingray were collected from 
an area of small-scale, shallow-water shrimp fishing oper¬ 
ations in the central zone of the Pacific coast of Colom¬ 
bia (from 4°34'N, 77°21'W to 2°31'N, 78°34'W), during 
2006-2009 and 2015. The number of animals collected 
was lower during the first quarter of the year, given the 
closure of the shrimp trawl fishery in these months along 
the Pacific coast of Colombia. This area has sandy-muddy 
substrate and shallow (<8.3-m depth), warm (between 
25°C and 29°C), and brackish (salinity between 21.8 and 
25.6) waters (Mejfa-Falla, 2012). 
Rays were measured (disc width [DW], in centimeters), 
their sex was determined, they were eviscerated, and part 
of the vertebral column from the abdominal region was 
extracted, tagged, and frozen until analysis was done later 
in the laboratory. Differences in DW between sexes were 
evaluated by using a Mann-Whitney test. 
Vertebrae processing and band-pair counts 
Vertebrae were cleaned manually by using a scalpel to 
eliminate the excess of muscle and connective tissue 
and to obtain the vertebral centra. The diameter of ver¬ 
tebrae was measured sagittally, and then the vertebrae 
were grouped in 3 size categories: large (2.50-3.30 mm), 
medium (1.60-2.49 mm), and small (0.70-1.59 mm). 
Tests were carried out by using different thicknesses of 
cut, stains, and times of staining for each vertebra size. 
Vertebral sections were rinsed with distilled water and 
polished before being observed. The best combination for 
the visualization of bands was 0.4-mm sagittal sections, 
stained with light green (0.05%) for 5 min for small ver¬ 
tebrae and with methylene blue (0.001%) for 10 min for 
medium vertebrae and for 20 min for large vertebrae. 
Sections were observed by using an optical microscope 
under transmitted light. Photographs were taken along 
with the respective scale by using the Zen lite 1 (blue ed., 
vers. 1.0) microscope software for light microscopy sys¬ 
tems (Carl Zeiss Microscopy, LLC, Jena, Germany), and 
the growth bands were measured along the border of the 
corpus calcareum of each vertebra, by using the Zen lite 
microscope software. 
Two readers (reader 1 had the most experience) per¬ 
formed a training exercise of counting the bands of a 
subsample (sample size [ft]=100) to refine the methods, 
identification criteria, and counts of translucent bands. 
The following criteria were established: 1) identification 
of the presence of a pair of growth bands (one translucent 
and one opaque), defined as a band pair; 2) identification 
of a birth band through a change in the angle of the corpus 
calcareum in the closest place to the focus of the vertebra; 
and 3) enumeration of translucent bands. The 2 readers 
then did an independent band count of the whole sample, 
without knowing the sex or size of the specimens. The 
translucent-band counts were compared, and if the band 
counts did not coincide between readings (with differences 
in band counts of 3 or more), a third reading was made to 
reach a consensus; if no consensus was reached, the sam¬ 
ple was discarded. 
Bias and precision between readers 
Different methods were employed to evaluate bias and 
precision of readings, on the basis of independent readings 
(Cailliet and Goldman, 2004). The bias between readers 
was evaluated through an age-bias plot (Campana et al., 
1995) and Bowker’s test for symmetry (Hoenig et al., 
1995). The age-bias plot relates the band-pair count of 
reader 1 (x-axis) versus the mean band-pair count of reader 2 
(y-axis). The second test was used to evaluate whether dif¬ 
ferences between readings by reader 1 and 2 were due to 
random events (P>0.05) or systematic errors (P<0.05), by 
using a chi-square test. 
The percentage of vertebrae that could be read was 
calculated as (number of read vertebrae/total number of 
samples)xl00. The percentage of agreement between 
readers was calculated as (number of agreements/total 
number of readings)xl00, taking into account differences 
of 0, 1, and 2 band pairs. 
The average percent error (APE) and the coefficient 
of variation (CV) between readers were also evalu¬ 
ated, taking into account variations of each reading 
with respect to the mean value of readings, as follows 
(Beamish and Fournier, 1981; Campana et al., 1995; 
Campana, 2001): 
APE 
■ 
( 
l l Y 
1 
r* h- x i| 
ft 
R l 
_ 
\ 
3 /J 
x100 and 
(1) 
1 Mention of trade names or commercial companies is for identi¬ 
fication purposes only and does not imply endorsement by the 
National Marine Fisheries Service, NOAA. 
