Molina et al.: Age and growth of Mustelus schmitti 
367 
Figure 1 
Map of the study area showing the 3 sampling stations, Los Pocitos, 
Ria, and San Bias, where narrownose smooth-hound ( Mustelus schmitti ) 
were captured in 2008 in a protected area, the Reserva Natural de Uso 
Multiple Bahia San Bias, located in Anegada Bay, Buenos Aires Province, 
Argentina. 
stripped of soft tissue with a scalpel and 
frozen for transportation and storage. 
Preparation and sectioning of vertebrae 
Vertebrae were prepared by following 
the methods described in Cailliet et al. 
(1990), Moulton et al. (1992), Natanson 
et al. (1995), Campana (2001), Conrath 
et al. (2002), and Lessa et al. (2016). 
Samples were cleaned of excess tissue 
and separated into individual centra so 
that they could be submerged in a so- 
lution of 2.5% sodium hypochlorite to 
remove connective tissue, without com- 
promising the interpretability of growth 
bands. Bath duration varied between 5 
and 30 min, depending on the size of the 
centra. Afterward, vertebrae were rinsed 
with tap water, air dried, and mounted 
on transparent epoxy resin. One section 
(approximately 0.2 mm thick) was then 
cut transversely through the focus with a 
diamond-bladed IsoMet Low Speed Saw 7 
(Bueheler, Lake Bluff, IL). 
Age determination 
Vertebral sections were examined under 
a binocular microscope with transmitted 
light to identify opaque (hypomineralized, 
wide) and translucent (hypermineralized, 
narrow) bands. All counts were made 
with no knowledge of size, sex, or date 
of capture of the individual. The birth 
mark was identified by the angle change 
in growth bands and by the banding pat- 
terns on neonates and embryos. Bands consisting of 
opaque and translucent areas were identified along 
the corpus calcareum (Goldman et al., 2012) (Fig. 2) in 
each vertebra. Variation in the nature of the border of 
the vertebra was used to verify the temporal periodici- 
ty of the formation of each band type (Campana, 2001). 
The age of each individual fish was determined by the 
number of translucent bands, the date of capture, and 
the date of birth, which was assumed to be 1 January. 
This assumption is supported by the fact that a high 
proportion of gravid females were captured in a previ- 
ous sampling campaign in December 2007 and by the 
high number of young of the year captured in February 
2008 (Colautti et al., 2010). For example, a fish that 
was captured on 12 February 2008 and had 2 bands 
was calculated to have an age of 408 d (1.12 years). 
Bias and precision of age estimations were assessed 
according to the methods proposed by Goldman et al. 
(2012). The thin sections of vertebrae from the same 
7 Mention of trade names or commercial companies is for iden- 
tification purposes only and does not imply endorsement by 
the National Marine Fisheries Service, NOAA. 
specimen were read twice by 2 different readers, and 
the analysis of bias in age determination was carried 
out with a bias plot (Campana et al., 1995; Officer et al., 
1996). The indices of average percent error (Beamish 
and Fournier, 1981) and the average coefficient of 
variation (Chang, 1982) were calculated to assess the 
precision of the age determination between the 2 read- 
ers as well as between both readings of each reader 
(Ogle, 2015). Differences in readings between readers 
were reconciled by both readers reading a third time 
the samples for which they had different band counts. 
If agreement on a band count was not reached with the 
additional readings, the sample was eliminated from 
further analysis. 
Estimates of somatic growth 
To model growth, we used partial ages because they 
have been shown to be more accurate than rounded 
ages (Smart et al., 2013). To calculate the partial ages, 
we used the fraction of the year that had elapsed from 
the time of birth to the time each individual was col- 
lected. Partial ages were then grouped in 0.5-year bins. 
