Karlou-Riga and Economidis Ovarian atretic rates and sexual maturity of Trachaurus trachurus 7 r > 



Oct Dec 

 1989 



Month and year 



Figure 6 



Gonadosomatic (mean GSI) and hepatosomatic (mean HSI) index of 

 European horse mackerel, Trachurus trachurus, females per sam- 

 pling month during the period from October 1989 to May 1991. Num- 

 ber of females examined = 980. Solid line = GSI; dashed line = HSI. 



son as was indicated by high atretic states. During the 

 second reproductive period, we noticed that GSI attained 



^ — i — i — \ — i I i  i — \ — i — i — \ — i — \ — \ — [- 



155 175 195 215 235 255 275 295 315 335 355 375 



Total fish length (mm) 



Figure 7 



Fraction of European horse mackerel, Trachurus trachurus, fe- 

 males that were sexually mature on the basis of histological cri- 

 teria (crosses and solid line) and macroscopic criteria (filled 

 squares and dashed line) during two successive reproductive pe- 

 riods. The ML 50 , based on 307 mature females, was estimated 

 histologically to be 211 mm (logistic curve parameters: a = 17.104, 

 6=0.081, r 2 =0.98); the ML 50 based on 595 mature females, was 

 estimated macroscopically to be 220 mm (logistic curve param- 

 eters: a = 17.863, 6=0.081, r 2 =0.99). 



about one third of the value recorded in the 

 previous period. The lower GSI values during 

 the second period also support the high num- 

 bers of postspawning females that were ob- 

 served histologically during mid-season, result- 

 ing in a possible drop of annual fecundity. 



Sexual maturity 



We found that the values of ML g0 estimated 

 histologically (211 mm) and macroscopically 

 (220 mm) were very similar. This indicates 

 that the separation between mature and im- 

 mature females by gross examination of the 

 gonads was relatively unbiased. However, a 

 number of females with early developing or 

 regressing ovaries were grossly misclassified 

 as immature. These could be correctly iden- 

 tified only by histological examination, re- 

 sulting in lower proportions of mature fe- 

 males with the macroscopic method. On the 

 other hand, it is possible that ML 50 , deter- 

 mined histologically, may be overestimated 

 when measurements are made during the 

 spawning season (Hunter et al., 1992). This 

 did not happen in the present work because im- 

 mature females were identified only at the be- 

 ginning of the reproductive period, before 

 spawning had started. Consequently, we believe 

 the ML 5(1 determined from histological sections 

 was the least-biased estimate. Although the age 

 at first maturity (i.e. 2 years) is in general agree- 

 ment with Arruda's ( 1984 ) estimation, our ML 50 

 values were not strictly comparable to those of 

 Arruda, because the latter refers to standard 

 length units and no conversion equation is 

 given. However, our value was probably lower 

 because Arruda considered only macroscopically 

 "ripe" individuals as mature. A smaller num- 

 ber of stages used to define maturity was also 

 the reason for a higher value of ML 50 (254 mm) 

 estimated by Kerstan (1985). Finally, the dif- 

 ferences in sexual maturity referred to by vari- 

 ous investigators should be compared carefully 

 owing to possible different sampling times or, 

 more importantly, to a strict use of anatomical 

 criteria to identify the mature individuals re- 

 sulting in a significant bias, especially for mul- 

 tiple spawners. 



Acknowledgments 



We thank J. Hunter for his comments about 

 atresia and postovulatory follicles, B. Macewicz 



