Environ Biol Fish (2009) 84:79–87 DOI 10.1007/s10641-008-9391-1
Validation of annulus in otolith and estimation of growth rate for Japanese eel Anguilla japonica in tropical southern Taiwan Yu-Jia Lin & Wann-Nian Tzeng
Received: 8 January 2008 / Accepted: 24 July 2008 / Published online: 30 August 2008 # Springer Science + Business Media B.V. 2008
Abstract The age of Japanese eels (Anguilla japonica) is often estimated from otoliths, but this method has not been fully validated, particularly in tropical areas where the annulus in otolith is considered to be less distinct than in temperate areas. To validate the annuli in Japanese eel otoliths from southern Taiwan, knownage (2 year-old) cultured eels from an eel farm and wild eels from Kao-Ping River were collected. It was found that 26 out of 31 cultured eels (83.9%) showed two clear annuli and the remained 5 eels showed either one or three annuli. The mean (± SD) age of the cultured eels was 1.97±0.4 years. Meanwhile, a clear peak in the mean monthly marginal increment ratio of the otolith in wild yellow and silver eels occurred once a year during winter (November to March). The annual deposition of presumed annuli in otoliths of Japanese eel was validated and the age and growth rate estimation for Japanese eels in the tropical southern Taiwan is deemed feasible. The growth rate of cultured eels was significantly faster than that of wild eels, but it did not Y.-J. Lin : W.-N. Tzeng Institute of Fishery Science, College of Life Science, National Taiwan University, Taipei 10617 Taiwan, ROC W.-N. Tzeng (*) Department of Life Science, College of Life Science, National Taiwan University, Taipei 10617 Taiwan, ROC e-mail:
[email protected]
differ significantly between sexes for wild silver, yellow or cultured eels. The von Bertalanffy Growth Function parameters (K, L1 and t0) of the wild eels were estimated as 0.114±0.028 year−1, 1178±171 mm and −0.8± 0.2 years, respectively. Keywords Japanese eel . Age validation . Growth . von Bertalanffy growth parameters
Introduction Accuracy and precision in age estimation is very important in understanding the life history traits and population dynamics of fish. Some population parameters such as growth rate, mortality rate and time at maturity depend strongly on accurate age determination (Quinn and Deriso 1999). Among the numerous hard structures used for age estimation (i.e. scale, otolith, vertebrae, etc.), the otolith is generally considered to be the best for estimating the age of many teleost fishes (Campana and Thorrold 2001). However, annulus formation in otoliths requires validation before it can be used to estimate the age of the fish (Beamish and McFarlane 1983; Campana, 2001). Japanese eel (Anguilla japonica) is a catadromous fish, widely distributed in Taiwan, China, Japan, and Korea (Tesch 2003). The eel spawns in the tropical Pacific Ocean west of the Mariana Islands (Tsukamoto 2006). The hatched larvae, leptocephali, are trans-
DO09391; No of Pages
80
ported passively via both the North Equatorial current and Kuroshio and metamorphose to glass eels on the continental shelf and further develop into pigmented elvers in the estuary (Cheng and Tzeng 1996). The eels were found to be facultatively catadromous, i.e., instead of migrating upstream to colonize freshwater habitats, a part of elvers stay in brackish estuaries or seawaters until sex maturation (Tsukamoto and Arai 2001; Tzeng et al. 2002). After 4–10 years, the yellow eels begin sexual maturation to become silver eels (Han et al. 2001, 2003), and then migrate to the spawning ground to spawn and die (Tesch 2003). Annulus formation in otoliths has been validated in some anguillid species in the temperate areas (Berg 1985; Oliveira 1996; Graynoth 1999; Pease et al. 2003). Although the annuli in otoliths of the Japanese eel Anguilla japonica are assumed to be deposited annually in studies of the eel age and growth (Tzeng et al. 2002), this has not yet been completely validated (Guan et al. 1994), particularly in tropical areas. It was hypothesized that the formation of annuli in otoliths of Japanese eels in the tropical Kao-Ping River in southern Taiwan followed an annual rhythm, similar to other anguillids. However, the annual deposition rhythm in the otoliths of tropical fishes has been considered problematic because of less seasonal variation in water temperature and relatively stable growth of the fish. This probably introduces controversy in the age and growth estimation of the eels in the tropics (Morales-Nin 1989; Morales-Nin and Panfili 2005). Moreover, the supernumerary (false or incomplete) growth rings found in some eel species are found in Japanese eels (Berg 1985; Vøllestad 1985; Michaud et al. 1988; Svedäng et al. 1998; Graynoth 1999; Pease et al. 2003), which further stresses the fact that the annulus validation for the Japanese eel is particularly important to exclude the ambiguity in age and growth rate estimation. This study aims to validate annulus formation in the otoliths of Japanese eels in the tropical southern Taiwan by the known-age cultured eels collected from eel farms in Donggang Town and by marginal increment analysis using wild eels collected from the lower reach of Kao-Ping River. Additionally, the growth rates of the eels were calculated and compared between cultured and wild eels and between sexes. The growth curve was fitted by von Bertalanffy growth equation.
Environ Biol Fish (2009) 84:79–87
Materials and methods Sampling of wild and cultured eels Two groups of Japanese eels, wild and cultured, were used for annulus validation in this study. During 1999 to 2003, wild eels were collected from the lower reaches of the Kao-Ping River, the largest river in southern Taiwan, with a length of approximately 171 km and a drainage area of 3256 km2 (120°50’E, 22°40’N, Fig. 1). The mean water temperature data in the bottom of the river were compiled from Cheng (2005) and Cheng et al. (2007), which were measured by CTD from January 2001 to January 2004 and for the duration of 2006. The water temperature ranged from 21–23°C in the winter to 28°C in midsummer and was greater than 20°C all year round. Accordingly, the sampling area of the wild eels can be described as tropical (Pauly 1998). Cultured eels were collected randomly from the eel farms in Donggang Town on 26th March 2007 (22N° 28’, 120E° 26’, Fig. 1). These cultured eels had been reared from elvers since during the winter of 2004–2005, which implies that the cultured eels were approximately 2 years old after elver stage. The cultured eels were reared in the outdoor ponds with a depth less than 1.5 m without any temperature and light control and thus the water temperature in the ponds was closely related to air temperature. The mean seasonal air temperature in Donggang Town during 2005–2007, measured by Taiwan Central Weather Bureau, ranged from 18.5 to 20°C in winter (December to February) and 21 to 26°C in spring (March to May), reaching a maximum of about 28°C in summer (June to August) and then decreasing in autumn to 23–27°C (September to November). The cultured eels were fed generous amounts of commercial feed twice a day and thus the growth rate was quite fast. According to the eel farm owner, the size of the elver was about 50 mm in the beginning of the rearing. These eels all reached marketable size (larger than 350 mm) after approximately 26 months. Both wild and cultured eels were anesthetized immediately with ice after collection and then transferred to the laboratory and preserved at −20°C. After defrosting, the total length (TL, to nearest 1 mm) and total weight (TW, to nearest 0.1 g) were measured and sexes were determined by gross inspection of the
Environ Biol Fish (2009) 84:79–87
81
Fig. 1 Sampling sites of Japanese eel, Anguilla japonica. Wild yellow and silver eels were collected from the lower reach of Kao-Ping River (enclosed region) and the cultured eels were collected from the eel farms in Donggang Town
gonads. The developmental stages (silver or yellow) were identified based on body color, eye size, and blackness of the pectoral fins (Han et al. 2003; Tesch 2003). Yellow wild eels were randomly selected from three size classes (small, medium and large) to increase the sample size and more male and sexually-undifferentiated yellow eels were collected to balance the femalebiased sex ratio of the silver eels. Otolith preparation The sagittal otoliths were removed, cleaned, dried in the air, ground and polished until the primordium was exposed. Polished otoliths were then etched with 5% EDTA for 2 to 3 min to enhance the visibility of the annuli. The digital pictures of the annuli in otoliths were taken under the optic microscope with a reflected light and the age of the eel was estimated from the counts of annuli (Tzeng et al. 1994, 2002). The otolith and annular radii along the longest axis from the primordium to each annulus and to the otolith edge were measured to the nearest 1 µm using the image processing software (SigmaScan Pro 5.0, SPSS Inc.).
Calculation of marginal increment ratio The marginal increment ratio (MIR) of the otolith was calculated following Lessa et al. (2006): MIR ¼ ðRC RL Þ ðRL RL1 Þ1 where RC, RL and RL–1 is the otolith radius from the primordium to the edge, to the last annulus and to the annulus next to the last one. The mean MIR in each month was plotted to determine whether otolith annuli were deposited annually. Growth of the eels The linear growth rate (GR) was calculated to understand the somatic growth rate of the eels. GR represents the slop of the linear relationship between the age and total length of the eel, which was calculated as: GR ¼ ðTL L0 Þ Age1 where TL=the total length, L0 =the mean length of glass eels at estuarine arrival, which was 55.55 mm for
82
the eels in Kao-Ping River (Cheng and Tzeng 1996) and Age=the age estimated by the otolith readings. The growth curve of the eels after the elver stage was fitted by the von Bertalanffy Growth function (VBGF): h i Lt ¼ L1 * 1 eK ðtt0 Þ where Lt is the length in mm at time t, L1 is the asymptotic length, K is the Brody growth coefficient, and t0 is the initial condition parameter when the hypothetical length is zero (Quinn and Deriso 1999). To increase the sample size, the length-at-age at age i (LBi) was back-calculated using otolith radius at each age, in which two approaches, regression and proportion methods, were often applied. The regression methods were not used in this study for their low coefficient of determination (R2 =0.42 for log10-transformed data) compared to 0.98 in Jessop et al (2004) and 0.89 in Thibault et al. (2007), which might be due to the variation in grinding planes among otoliths. Meanwhile, proportional methods were suggested for incorporation of information on individual length and radii (Gutreuter 1987). Therefore, the proportional method, Dahl-Lea method (Francis 1990) was used: LBi ¼ Lc ðRi =Rc Þ where Lc is the length at capture, Ri is the radius from the primordium to the ith annulus and Rc is the radius from the primordium to the otolith edge. Statistical analysis The sex ratio between wild silver and captive eels was compared using the Chi-square homogeneity test. The yellow eels were excluded from the comparison of sex ratio because they were randomly selected from each sex with a pre-determined sample size. Because the growth rate (GR) of the eels was not normally distributed (Kolmogorov–Smirnov test, p
