Determination of age and growth of the striped seabream Lithognathus mormyrus ( Sparidae ) in the Canarian archipelago by otolith readings and backcalculation *

The striped seabream Lithognathus mormyrus (Linnaeus, 1758) is a marine fish belonging to the Sparidae family. It is a demersal species living in groups over various types of sea bottoms, especially rocky, sand and seagrass beds, at depths ranging from 0 to 150 m. This species is distributed in the eastern Atlantic and in the western Indian Ocean. In the eastern Atlantic, it occurs from the Bay of Biscay to the Cape of Good Hope, and around the Canaries and Cape Verde. It is also present in the Mediterranean, Black, Azov and Red seas. In the western Indian Ocean, it occurs on the Natal coast (Smith and Smith, 1986; Bauchot and Hureau, 1990; Harmelin-Vivien et al., 1995). In the Canary Islands, the striped seabream is one of the main target species of the demersal smallscale fishery (Pajuelo, 1997). This species is caught at depths ranging between 10 and 100 m with traps. It is captured all year round with seasonal differences in landings. Catches of striped seabream have been declining in this area during the last few years. Despite its fishing importance, the striped seabream has never been the object of investigation in the Canary Islands. The present work investigated SCI. MAR., 66 (1): 27-32 SCIENTIA MARINA 2002

In the Canary Islands, the striped seabream is one of the main target species of the demersal smallscale fishery (Pajuelo, 1997).This species is caught ai áepths ranging ' m e e n i0 anci iG0 m with mps.Zt is captured aii year round with seasonal differences in landings.Catches of striped s e a b ~a m have been deciining in this area during the last few years.
Despite its fishing importame, the striped seabream has never been the object of investigation in the Canary Islands.The present work investigated the age and growth of the striped seabream off the Canary Islands to obtain growth estimates, which are important input parameters for stock assessment techniques and will provide an insight into the lifehistory of this species.The importante of this study is enhanced by the fact that o d y three articles on the age and growth of the species have been published (Suau, 1970;KraljeviC et al., 1995KraljeviC et al., , 1996)).

MATERIALS AND METI-IODS
A total of 496 individuals of striped seabream were coiiected from commerciai catches of the artisanal fleet from January to December 1999.Fish were caught with traps deployed on the bottom at depths of 10-90 m off Gran Canana (Canary Islands).
For each fish, the total length (TL, m) was measured, the sex was determined (male, female or intersexual), and the sagittal otoliths were removed Age was determined by counting the a ~u a l growth rings on the otoliths.Whole otoliths were placed in glycerin and examined under a compound microscope (15 x) with reflected light against a dark background.Counts for each otolith were performed three times.Readings for given otoliths were accepted only when two agreed An index of average percent error was used to compare the accuracy of the deterrnination (Beamish and Foumier, 1981).Ageing was validated indirectly by means of the analysis of the evolution of the mean monthly marginal increments.The marginal increment (MI, 0.01 m ) was measured as the distance from the outer mar& of the outermost translucent ring to the periphery of the otolith.Measurements were always made along the longest axis of the piece.Once the annuai periodicity of the rings was confinned, individuals were assigned to the corresponding age classes beginning with the date of capture and considering 1 September as the b i d a t e .The von Bertalanm growth curve was fitted to the data of the resulting agelength key by means of Marquardt's algorithrn for non-linear least squares parameter estirnation (Saila et al., 1988).
The backcalculation method was applied as an independent means of validating the otolith-based age detenninations (Morales-Nin, 1989).Backcalculation analysis was undertaken using a method described by Francis (1990):The radius of the ith band (RI, 0.01 mm), the distance from the centre of the otolith to the outer edge of the translucent ring, and the radius of the otolith at capture (R, 0.01 mm), i.e. the distance from the centre of the otolith to the periphery, were measured.Measurernents were always made along the longest axis of the otolith.The size of an individual when the ith band was laid down (SI, mm) was calculated as SI=(RVR)bTL, where b is a constant derived from the power function which describes the relationship between the radius of the otolith and the total length of the fish (Francis, 1990).The von Bertalanffy growth curve was fitted to the backcalculated mean length at age by means of Marquardt's algorithm for non-linear least squares parameter estimation (Saila et al-,  1988).The growth parameters obtained for males, females and ali individuais were cornpared statistically by means of Hoteiling's T2-test (Bemard, 1981).
To compare the growth pararneters obtained in the present papers with those reported by other authors for the same species, the growth performance index (8) was used (Pauly, 1991).This index has the fom: 8 = 2 log,,L-+ log,,,k, where k is the growth coefficient, and L_ is the asymptotic length.A concentric pattem of translucent and opaque zona was readily dintinguishable in the otoliths, and easily interpreted.Of the 496 otoliths examined, 17 (3.4%)were considered unreadable and therefore no age estimates were obtained from thern.Of these otoliths, 6 were broken, and 11 had poorly defined growth zones.Of the remaining 479 otoliths, the readings were coincident at least twice in 452 (94.5%) but yielded conflicting ages in 27 (5.5%).The value of the index of average percent error was only 3.1%.Annual marks were equal in both the anterior and lateral otolith fields.Readings were equal in both left and ngth otoliths.The evolution of the mean monthly marginal increments in oíoiiins wiüi one aW iwo ñansiuceñí zones, three and four iranslucent zones, and more than four translucent zones is shown in Figure 1.The same temporal variation pattem was recorded for the three cases.The highest values occurred between May and October, peaking during July-August.From November to April the values were low.Based on this, it was assumed that an opaque zone and its adjacent translucent zone were deposited on the otoliths each year.The values of increments on the otoliths decreased wiih the age of the fish.
Fish aged O to 8 year old were recorded (Table 1).Individuals attained over 50% of their maximum observed size during the second year of life.Growth parameters determined for males, females, and all individuals (males, females, immatures and intersexuals) are shown in Table 2. Signifícant differences in the growth parameters were found between males and females (Hotelling's F-test, Fish total length and otolith radius were closely correlated (Fig. 2).The total length and otolith radius were also estimated for males (LT= -Age-length key for aU fish of miped seabream L mormyms off the C a n q blandr AGE AND GROWTH OF STRiPED SEABREAM 29 37.216RI3") and females (LT=38.022R1397)separately.Backcalcuiated total lengths at ihe end of ea& year of life by age grwp for al1 individuals are shown in Table 3.There was no indication of Rosa Lee's phenomenon in which compted sizes at a @ven age tend to be smaller when derived from measurements on older fish (Francis, 1990).Observed lengths were slightiy higher than backcalculated lengtbs for individual age groups.Growth parameters estimated from the mean backcalculated sizes at age for males, females, and d i individuals are presented in Table 2. Significant differences in &e growth parameters were found -between males No significant differences óetween the growth paramaers estimated by reading otoliths and using  lo and Lorenzo, 1995lo and Lorenzo, , 1996lo and Lorenzo, ,1998lo and Lorenzo, , 1999lo and Lorenzo, ,2000)).
The proportionality between fish growtfi and otolith size increase allows backcalculation to be used to determine the growth.The results obtained by backcaiculation were very satisfactory and confirmed the estimations of the age and growth of the striped seabream of the Canary Islands by otolith readings.Because the nng formation is regular and therefore the otoliths can be used for age determination, and because the fish length and otolith size are closdy correlated, it is judged valid to p&t the use of measurements to previously formed marks to backcalculate the growth history Wagenal and Tesch, 1978;Barlett et al., 1984;Campana, 1990;Francis, 1990).
The striped seabream off the Canary Islands has a relatively short life (8 years).Kraljevid et al. (1995,1996) also found that this species may attain an age of 8 years in the Adriatic Sea, and Suau (1970) exarnined 7-year-old specimens from the Spanish Mediterranean coast.
In the Canarian archipelago, the striped seabream grows relatively fast during the first few years of life, attaining approximately fifty percent of its maximum length during the second year.M e r &e second year, tiie annual growth rate drops rapidly, related to sexual maturity, since in the studied area indi-vidual~ are mature by the second year of life (unpublished data).Hence, energy seerns to be diverted to reproduction (June-Decernber), with less energy available for somatic growth.The difference in giowth between sexes, with females reaching a slightly larger len@ than males at the sane age, is a characteristic of the protandric qwxies such as the striped seabream ( Besseau andBmslBSicard, 1991, 1995).Alekseev (1983) and Shapiro (1984) pointed out that the mean SUR of sex-reversed fish is siightly greater than that of males of the same age.Therefore, the difference in size between males and femaies of the same age cannot be considered as evidence of an intersexual diñerence in growth rates because d e s and females are the same specimens at áiEerent stages of sexual succession and h e largest males in an age group may be tbe first to revert.
The growth parameten of the striped seabream off the Canary Islands are very similar to those obtained by Suau (1970) and KraljeviL et aL (1995,1996) for L mnnyrus on the Spanish Meditaranean coast (L_=332 m, ka.275 y-') and in the Adnatic sea (L_=362 cm, k=0.297 yr-') respectively.The growth performance index obtained for the striped seabream of the Canary Islands (fk4.53) is v e q sirniiar to that estimateti from the growth parameters given by Suau (1970) on the Spanish Mediterranean coast (0=4.48) and Kraljevid et al.  (1995, 1996) in the Adriatic Sea (8=4.59).
Of the 496 individuals exarnined, 235 (47.4%) were males, 229 (46.2%) females, and 24 (4.8%) intersexuals.The sex of the remaining 8 (1 -6%) indi-vidual~ could not be identified macroscopically because they were imrnature and had very thin, translucent gonads.The size range of the individuals was between 113 and 350 mm total length.Males ranged in size from 201 to 350 mm, females from 199 to 350 mm, and intersexuals from 217 to 323 mrn.The length of immature fish was compised between 113 and 230 m.

FIG. 1 .
FIG. 1. -Mean montfity evolution of dx marginal increment in the otolitbs of mi@ seabrcam L monnyms off the Canary island..
. 1 ~ Tg20.M = I .o Y ) .rne mean size oi males was slightly smaller than the mean size of females at the same age.