852 



Fishery Bulletin 101(4) 



Figure 1 



Saggital otoliths of P. incisus collected off the Canary Islands: (A) from a two-year-old (182 mm); (B) from a 

 three-year-old (202 mm); and (C) from a five-year-old (265 mm) fish. Ri is the radius of the i"' band (distance 

 from the centre of the otolith to the outer margin of the annulus), Re is the radius of the otolith at capture, and 

 MI is the marginal increment measured. 



by two readers without knowledge of the size and sex of the 

 specimens, or previous counts of the other reader. Counts 

 were made for otoliths of each individual on two separate 

 occasions, and only coincident readings were accepted. The 

 same approach was used to determine the final number 

 of bands in each specimen, and a consensus was reached 

 between readers on the final counts. Reproducibility of the 

 resultant age estimates was evaluated with the coefficient 

 of variation (CV) (Chang, 1982) and the index of average 

 percent error (lAPE) (Beamish and Fournier, 1981). 



To validate that rings were formed annually in the 

 bastard grunt, we analyzed the monthly mean marginal 

 increment (Hyndes and Potter, 1997). Marginal increment, 

 estimated as the distance between the outer edge of the 

 outermost annual ring and the periphery of each otolith, 

 was measured (to the nearest 0.01 mm) with an ocular 

 micrometer (Fig. 1). Measurements were always made 

 along the longest axis of the otolith. The pattern expected 

 in the marginal increment would be a minimal value at the 

 start of the growth period, increasing with time until the 

 measurement fell to a minimum again at the formation of 

 the next period of growth (Pilling et al., 2000). The size of 

 the marginal increment varies both with the age of the fish 

 and the time of sampling during the year. Because older 

 fish grow slower than younger fish, a smaller marginal 

 increment is expected. For this reason, to assess the pos- 

 sibility of false annulus formation among either younger 

 or older bastard grunt, quantitative marginal increment 

 analyses should be standardized for age. Therefore, we 

 used age class to standardize our analyses. Owing to the 



wide range of ages encountered, however, there were insuf- 

 ficient samples to fully accomplish this standardization. 

 It was necessary to combine the ages in two or more age 

 groups representing fast, moderate, and slow-growing in- 

 dividuals (Pilling et al., 2000). Mean marginal increments 

 were plotted against month of capture, and the minimum 

 was used to indicate the month of annulus formation. 



Once the periodicity and timing of ring formation were 

 verified, the age of each fish was determined from the num- 

 ber of annuli, the assumed birthdate, and the sampling date. 

 It was assumed that annulus formation began 1 January, 

 corresponding to the peak of spawning in the species 

 (Gregoire'). The difference between the date of capture and 

 the birthdate was used to estimate a fractional age (Gordoa 

 and Moll, 1997). This fraction was added to the number of 

 annuli read in the otoliths to avoid any potential bias 

 in growth estimates due to differences in sampling date 

 (Gordoa and Moli, 1997). 



Length-at-age was described by the three-parameter 

 specialized von Bertalanffy growth model (Ricker, 1973): 



L, =LJl-e 



-kH-l„) 



), 



the four-parameter Schnute growth model (Schnute, 

 1981): 



Gregoire, M. 2001. Edad y crecimiento del roncador 

 Pomadasys incisus (Bowdich, 1825) de Gran Canaria (Islas 

 Canarias). Unpubl.data. ULPGC Socrates/Erasmus Research 

 Report, 34 p. Departament of Biology, University of Las Palmas 

 de Gran Canaria. 35017 Las Palmas de Gran Canaria. Spain. 



