Kwok: Age and growth of Tnchiurus spp 



749 



ventral 



mm) of T! nanhaiensis were obtained from commercial 

 catches in the coastal waters of Hong Kong. Commer- 

 cial gears included longlines, purse seines and bottom 

 trawls. Fresh specimens were placed on ice, trans- 

 ported to the laboratory, and identified by using the 

 diagnostic key of Wang et al. ( 1992. 1993): if the frontal 

 bone split laterally, specimens were identified as T. lep- 

 turus. otherwise T. nanhaiensis. Preanal lengths were 

 measured to the nearest mm. Specimens were blotted 

 dry and weighed (whole and gutted) to the nearest 

 0.01 g. To estimate the relationship between PL and 

 gutted weight (W), the variables were log-transformed 

 to meet the assumptions of normality and homoge- 

 neous variance. A linear version of the power function: 

 W (g) = a PL'' (mm) was fitted to the data. 



Distinct growth rings on whole otoliths and verte- 

 bral centra were ill-defined. Transverse sections of 

 sagittal otolith yielded "readable" growth rings; the 

 latter were chosen as aging tools. Left and right sag- 

 ittae were weighed independently to the nearest 0.01 

 mg after being oven dried at 40°C for 

 30 min. Otolith length (OL) was mea- 

 sured to the nearest 0.05 mm with cali- 

 pers. Sagittal otoliths were embedded in 

 resin and sectioned transversely through 

 the nucleus with a low-speed saw. Up to 

 five sections, 0.3 to 0.5 mm thick, were 

 made from each otolith to ensure that at 

 least one passed through the center of 

 the nucleus. Sections were then ground 

 with 1000- and 1200-grit sand paper, 

 mounted on glass slides with clear fin- 

 gernail polish, and examined with a com- 

 pound microscope at 40x magnification 

 with transmitted light. The relative age 

 in years was determined by counting the 

 number of opaque growth rings on the 

 dorsal side of the sectioned otoliths (Fig. 

 2). Thirty-five pairs of sectioned otoliths 

 of both species were processed. No dif- 

 ferences in the number of growth rings 

 were found in left and right sections of 

 each pair. Thereafter, the right sagitta 

 was used for age determination. 



Otoliths were read twice (one month 

 apart) in a random order, with no knowl- 

 edge of fish length or species. Precision was measured by 

 the percentage of agreement between readings (Lowerre- 

 Barbieri et al., 1994). Deviations were counted a third 

 time. Only counts with at least two agreements were used 

 in subsequent analyses. Marginal increment method was 

 used to validate the reading of annuli. Otolith radius (OR), 

 otolith annular radius (OAR), and marginal increment 

 (MI) (Fig. 2) were measured with an ocular micrometer to 

 the nearest 0.025 mm. 



The tendency for older fish to reflect smaller back-cal- 

 culated length at earlier ages than measured length is 

 known as Lee's phenomenon (Smith, 1983), and is related 

 to size-selective mortality (Boehlert et al., 1989). To eval- 

 uate this phenomenon, the mean otolith annular radius 



25°N 



20°N 



110"E 



IIS^E 



120°E 



Figure 1 



The South China Sea and surrounding area. 



sulcal 

 groove 



medial 



Figure 2 



Schematic diagram of a transversely sectioned sagittal otolith of cutlassfishes, 

 Trichiurus spp. MI = marginal increment, OARl = first otolith annular radius, 

 0AR2 = second otolith annular radius, OR = otolith radius. 



(MOAR) for each annulus of the same age group was 

 calculated, and the MOAR for each annulus of different age 

 groups was plotted against the age group (Yamaguchi et 

 al., 1990). Thus, we determined if older fish demonstrated 

 slower growth of hard parts at younger ages, i.e. true 

 Lee's phenomenon (Smale and Taylor, 1987). The ANOVA 

 test was used to compare MOARs among different age 

 groups. 



Multiple linear-regression models were fitted in a step- 

 wise manner to predict age from otolith weight and length. 

 Variables were log-transformed to meet the assumptions 

 of normality and homogeneous variance. A paired-sample 

 f-test showed no significant difference between left and 

 right sagittal otoliths in terms of weight (T. lepturus: 



