Benthos Research, 43 : 53-66, Jul., 1992 Age Determination and Growth Analysis Based on External Shell Rings of the Protobranch Bivalve Yoldia notabilis YOKOYAMAin Otsuchi Bay, Northeastern Japan MASAHIRO NAKAOKA Ocean Research Institute, University of Tokyo 大 槌 湾 に生 息 す る ブ リソデ ガ イYoldia notabilis YOKOYAMAの 外 部 成 長 輪 に よ る年 齢 査 定 お よ び成 長 の解 析 仲 岡 雅 裕 Abstract NAKAOKA, MASAHIRO(Ocean Research Institute, University of Tokyo). 1992. Age Determina- tion and Growth Analysis Based on External Shell Rings of the Protobranch Bivalve Yoldia notabilis YOKOYAMAin Otsuchi Bay, Northeastern Japan. Benthos Research, 43: 53-66. Age structure and growth pattern of Yoldia notabilis in Otsuchi Bay, northeastern Japan were studied using samples collected monthly during the period between December 1989 and December 1990. Y. notabilis has growth rings on external shell surface, and it was found that the formation of the "major ring" occurs annually during late winter to early spring. Age of Y. notabilis can be determined up to seven years old by counting number of the major rings, whereas only four younger year classes are separable from size-frequency histograms. Mean shell lengths of the four youngest year classes separated by the two methods are identical to each other. The growth rate of Y. notabilis is very slow at 0 year old ; achieving only 1.3mm in shell length by the first winter. The growth becomes rapid after the first year and the animal reaches 32.0mm at the age of seven years old. The obtained growth curve shows a sigmoidal pattern and the Gompertz growth equation gives a best fit for the data. age structure by extracting cohorts from size- Introduction frequency histograms has been undertaken widely Age determination is essential in studies of in various benthic invertebrates (see TSUTSUMI & population dynamics of long-lived species which TANAKA, 1987). However, this method often contain size overlapping year classes. Analysis of requires large number of samples to distinguish year classes. Moreover, information attained using Received August 23, 1991: Accepted March 26, this method indicates only the characters of the 1992. population at the sampling time, but not of its 53 No.43 Benthos Research Jul., 1992 history. population by external ring measurements. Fur- For animals which possess accretionary grow- ther, the results are cross-checked by the usual ing body parts such as molluscs and echinoderms, it "size -frequency histogram" method in order to is possible in some cases to determine their age demonstrate that the ring can be accurately used as using external or internal growth rings. Age deter- an age marker. mination using these rings has an advantage over Materials and Methods those using size-frequency histograms because the age of animals can be determined individually. It is The samplings were carried out monthly also possible to examine the past growth history of between December 1989 and December 1990 at a each individual using the rings (SEED,1980). For permanent sampling station (St. Y2) which was the application of this method, however, it is neces- established at the inner part of Otsuchi Bay (39•‹ sary to determine the period of the ring formation. 19.8'N; 141•‹54.8'E). The depth of the station is 12 Growth rings are produced in various cyclical pat- to 15m, and the bottom substratum consists of terns, (e.g. daily, lunar-monthly, or annually), or muddy sand. even with aperiodical stochastic events such as Four to nine replicate samples were taken each disturbance, disease and shell damage (see LUTZ& month using a 0.1m2 Smith-McIntyre grab sampler. RHOADS,1980; KENNISH,1980; for reviews). Sediment samples were sieved through a mesh of The analysis of growth using growth rings has 1 mm openings and living individuals of Yoldia been carried out for various long-lived lamelli- notabilis retained on the sieve were collected. branch bivalves including Macoma balthica (e.g. Because of their slender form of the shell, animals LAMMENS,1967; GREEN,1973; BEUKEMAet al., smaller than 1.6mm in shell length were washed 1977),Mercenaria mercenaria (e.g. KENNISH,1980; away through the sieve. In order to collect the PETERSONet al., 1983; JONES et al., 1990), Mya specimens in smaller size-fraction, two to five sub- arenaria (e.g. BROUSSEAU,1979; GOSHIMA,1982) samples of the area of 100cm2 were taken from and Mytilus edulis (e.g. THEISEN, 1973; LUTZ, grab samples each month during the period 1976; RODHOUSEet al., 1984). However, informa- between June, 1990 and December, 1990, and ani- tion on growth of protobranch bivalve Yoldia is mals retained between sieves of 0.5mm and 1 mm still limited in spite of its wide geographical distri- mesh openings were collected under a dissecting bution (COWAN,1968)and an important role in soft microscope. bottom communities (RHOADS,1963; RHOADS& In the laboratory, shell length was measured to YOUNG,1970; LEVINTON,1977).Only three studies the nearest 0.1mm with a caliper for animals larger have been published on the growth of Yoldia spp. than 2mm in shell length and to the nearest 0.02 hitherto ; SANDERS(1956) and LEWISet al. (1982) mm with a micrometer for those smaller than 2 for Y. limatula along the northeastern coast of mm. The number of the "major" rings on external North America, and HUTCHINGS& HAEDRICH shell surface (see results), if countable, was then (1984) for Y. thraciaeformis in the deep Atlantic recorded and shell length at each ring was mea- Ocean. sured individually. Yoldia notabilis is one of the major benthic Separation of year classes for individuals animals in Otsuchi Bay, northeastern Japan. It has retained on a 1mm-mesh sieve was carried out (1) growth rings on the external shell surface. Year by counting the number of the major rings, and (2) classes of the population of Y. notabilis in Otsuchi from size-frequency histograms by means of a Bay have been separated using size-frequency histo- nonlinear least-square method which was described grams (NAKAOKA,1989), but not by means of the and introduced into a BASIC program by AKAMINE growth rings. The present paper analyzes the age (1982). composition and growth pattern of the Y. notabilis Growth of Yoldia notabilis was analyzed by 54 Age and Growth of Yoldia notabilis measuring shell length at growth rings. In order to where B, C, b and c are constants. The parameters construct growth curve, mean shell lengths at suc- of these equations and correlation coefficient (r) cessive growth rings were fitted into three theoreti- which represents the degree of fit between empiri- cal growth equations; (1) the von Bertalanffy equa- cal and theoretical shell lengths were computed tion, using iterative nonlinear least-square regression analysis (SAS INSTITUTE, 1988). Lt=L•‡(1-e-k(t-to)) Results where L, is the shell length at time t, Lm the theoretical maximal shell length, k a growth con- 1. Observations of external growth ring stant, and to the theoretical time when shell length Dark, brown-colored rings are discriminated equals zero; from light, yellow-colored shell surface of Yoldia (2) the Gompertz equation, notabilis (Fig. 1). The color pattern is produced independently of the fine grooves which appear Lt=L•‡e-Bexp(-Ct) obliquely on the shell surface at narrow and regular and (3) the logistic equation, intervals. The rings are separable into two types according to the thickness in color and the distinct- Lt=L•‡/(1+be-ct) ness from the background. The thick, distinct rings Fig. 1 Yoldia notabilis. Photographs of shell surface showing external banding pattern. a: Shells possessing two types of rings; m=major ring, i- minor ring. Scale bar=5mm. b : Dorso-lateral view of a small shell showing the umbonal area ; p =prod issoconch, m=the smallest major ring. Scale bar =1mm. c : A large shell on which the number of the ring is uncountable near shell margin ; an arrow indicates the area where numerous rings are piled up and inseparable by eye. Scale bar-5mm. 55 Nn 43 Benthos Research Jul., 1992 (termed as "major ring") are formed at regular tainty, these larger animals were used only for the intervals, whereas the thin, faint rings ("minor measurements of the seventh and younger rings in ring") are seen irregularly between the major rings this study. (Fig. 1a). In the umbonal area, the smallest major ring is 2. Season of the ring formation recognized outside of prodissoconch (Fig. 1b). The Shell increment from the outermost major ring size of the prodissoconch is 200um in its longitudi- to shell margin was determined for animals with nal axis, and shell length at the smallest ring is less than eight major rings, and monthly mean around 1.3mm. increment was calculated for each of four size The number of the major rings is easily count- classes (0-10mm, 10-20mm, 20-30mm, and more able in the specimens with less than eight rings. In than 30mm, in shell length) separately (Fig. 2). In the specimens possessing more rings, however, it is all size classes, shell increment suddenly dropped in often difficult to count the exact number of the March, 1990, indicating the appearance of a new rings because the rings are piled up at the area near ring during February and March, 1990. The new to the shell margin (Fig. 1c). Due to this uncer- ring was discerned in 95% (35 out of 37) of the Fig. 2 Yoldia nolabilis. Seasonal changes in shell increment (mean +SD) from the outermost major ring to shell margin for each of four size classes. Solid squares represent shell increments from the outer- most ring to shell margin until February, 1990, and solid circles, after February, 1990.