A 
(op) 
A ok 
oo 
Reader 2 (no. of bands 
2) 
ze} 
c 
© 
a 
= 
) 
S 
& 
(op) 
— 
a) 
xe} 
© 
® 
oc 
Reader 3 (no. of bands) 
Fishery Bulletin 119(1) 
IAPE=8.76% | 
CV=12.48% 
D=79.12% 
7 
IAPE=8.35% | 
CV=11.81% 
D=77.11% 
IAPE=8.77% 
CV=12.04% 
D=78.71% 
8 
Reader 2 (no. of bands) 
Figure 3 
Age-bias plots of counts of growth bands on vertebrae of golden cownose rays (Rhinoptera stein- 
dachneri) sampled between 2008 and 2014 in the southern Gulf of California in Mexico. Black 
circles represent mean counts (A) of reader 2 relative to reader 1, (B) of reader 3 relative to reader 
1, and (C) of reader 3 relative to reader 2. Error bars represent the 95% confidence intervals for 
the mean number of band counts assigned by readers. The dashed line indicates 1:1 equivalence. 
Numerals on the superior axis represent the number of counts for each number of bands. Repro- 
ducibility and precision of age estimates were assessed by using the index of average percent error 
(IAPE), coefficient of variation (CV), and precision index (D). 
completion in September—October when the highest mean 
MI value was reached (Fig. 4A). We found strong evidence for 
a difference in posterior mean MI between bimester 2 
(March—April) and bimester 5 (September—October), a dif- 
ference that could serve as support for the annual pattern of 
growth-band formation (Suppl. Table 2) (online only). Results 
of the edge analysis indicate that opaque growth bands 
form during March—August (spring-summer) and trans- 
lucent bands form during September—February 
(autumn—winter), as expected for annual formation of 
growth bands (Fig. 4B). Therefore, the age structure of 
the golden cownose rays sampled in the study area 
