ABSTRACT: The conventional acetate peel method was modified to analyze the shell growth pattern of juvenile Manila clam Ruditapes philippinarum as small as 2 mm in shell length (SL). In the outer shell layer along the axis of maximum growth, two types of growth increments were observed: distinct increments and indistinct increments, which, respectively, do and do not continue to the middle shell layer. The distinct increments were found to be formed every two days in intertidal and shallow subtidal zones by field enclosure experiments of juveniles with datum points marked with alizarin complexone. Growth patterns of juveniles (12 mm SL) collected from the Seaside Park of Yokohama in Tokyo Bay were analyzed to confirm the modified method. Mean daily shell growth rate from April to July 2005 ranged 120–142 μm/day, which was reasonable as compared with previous studies. It was impossible to backcalculate the growth to the settlement size (i.e. 0.2 mm SL) because of erosion of the outer shell surface, and the smallest backcalculated minimum shell length was 0.8 mm. Fluctuations in daily growth rate were high, ranging 29–315 μm/day, and did not show a clear two-weekly rhythm.
Not only is the Manila clam Ruditapes philippinarum an important fisheries resource, it also plays a pivotal role in the nutrient cycle in shallow coastal waters by filter-feeding, as well as producing pseudofeces1 of suspended particulate organic matter. The fisheries production of the Manila clam has markedly declined for the last two decades in Japan. The annual production in Chiba Prefecture along Tokyo Bay, for instance, declined from 7 × 104 t at its maximum in the late 1960s to less than 2 × 104 t in 1979, mainly because of large-scale reclamation of the tidal flats in which clam fisheries had been intensively operated.2 However, clam production has continued to gradually decline to 8.6 × 103 t in 2004 even after the cessation of the reclamation. Although Toba2 suggested that the decline after 1980s was associated with the poor occurrence of juveniles based on periodic monitoring of stocking density, environmental factors causing the juvenile decline remain unknown. Production of the Manila clam in Kumamoto Prefecture, Ariake Bay, has also declined drastically since the late 1970s independently of reclamations,3 and similar cases are reported in many other areas in Japan. Ishii and Sekiguchi4 suggested that the production decline in Ariake Bay is caused by a shortage of larval supply in addition to a decrease in spawning stock size and an increase in juvenile mortality; however, as in the case in Tokyo Bay, environmental factors causing these problems are not well understood.
Elucidation of factors causing the decline of the Manila clam production has not been attained mainly because of insufficient background information on early ecologic characteristics, such as growth rate and mortality in the field. However, some cases are reported in which high mortality of juveniles smaller than 5 mm in shell length (SL) is thought to determine the stock size of the Manila clam.5,6 Thus, information on juvenile stages is indispensable for understanding the recruitment process.
Ecological study of the juvenile Manila clam is in its infancy partly due to methodological problems. Mortality estimation of the juvenile Manila clam in the wild is particularly difficult because of successive larval recruitment during the extended spawning season and highly variable individual growth rates within the same tidal flat7–10 that lead to the presence of similarly sized shells with different ages, limiting the accuracy of cohort analysis based on shell size frequency distributions.
Daily growth patterns of bivalve shells can be analyzed by preparing either thin sections or acetate peels (replicas) of shells. For convenience, acetate peel preparations are most commonly used, which involves grinding and polishing the shell section, etching with a dilute solution of acid, and making an acetate replica of the etched surface.11 Daily bands (increments) allow the age of an individual to be determined. The incremental nature of shell growth also provides a record of shell size-at-age using backcalculation techniques. Such estimates of shell size can then be used to examine the role of size and growth in regulating juvenile bivalve survival. However, daily growth analysis of the Manila clam has only been reported in larger individuals,12,13 and the methods used for the preparation of acetate peels for adult Manila clam13 are not suitable for small and fragile juvenile shells.
The present study modified the acetate peel method, referring to conventional fish otolith preparation methods so as to make it suitable for juvenile Manila clam shell handling, and increment formation periodicity was determined by a field enclosure experiment. The growth increment formation patterns of wild-caught juveniles were tentatively analyzed to confirm the validity of the modified method.
MATERIALS AND METHODS
Acetate peel preparation
Juvenile Manila clams of various sizes (∼1–10 mm SL) were captured in the Seaside Park of Yokohama (35°20′N, 139°38′E), Tokyo Bay and frozen so that the shells open when thawed. Body tissues were removed by gently scrubbing the shells with forceps. Dried juvenile shells were then embedded in epoxy resin (Epoxicure resin, Buehler, Lake Bluff, IL, USA) and sectioned along the axis of maximum growth (AMG)14 with a diamond cut-off wheel (ultra-precision multipurpose micro grinder, NSK Nakanishi, Tokyo, Japan), leaving small margin. The AMG is a line connecting the umbo to the ventral margin at the midpoint of the shell length (Fig. 1). The cross sections were then ground to the AMG on a lapidary wheel using 1000-grade grit followed by 2000-grade grit waterproof abrasive paper, and polished with wrapping film sheets (3M, St. Paul, MN, USA) using grits of 2000, 4000, 8000, 10 000 and 15 000 grades consecutively. Polished surfaces were examined by stereomicroscope for scratches using reflecting light, and polishing was repeated until all visible scratches were removed.
Combinations of HCl dilutions from 0.01 to 0.1 N and etching times from 5 to 60 s were tested for the best resolution for the juvenile shells. Since adhesion of CO2 bubbles produced by the reaction between HCl and CaCO3 on the polished surface may cause an unevenly etched surface, etching was performed in an ultrasonic bath to remove the bubbles. The etched shells were immediately rinsed in running tap water and air-dried.
Acetate replicas were made by flooding the etched surface with methyl acetate and applying a sheet of acetyl cellulose film (0.034 mm thickness, Bioden RFA replica film, Oken Shoji, Tokyo, Japan), which was gently pressed by a cotton-tipped swab and let dry. The prepared replicas were then compressed between two thin glass slides (0.55 mm thickness) to smooth out wrinkles and were observed under a phase-contrast microscope for increment count in the outer shell layer.
Growth increment formation periodicity
Growth increment formation periodicity in juvenile Manila clam was examined by field enclosure experiments for known periods. Prior to the enclosure rearing experiments, shell staining with alizarin complexone (ALC) and tetracycline were tested for use as a datum point marker. Juvenile Manila clams were immersed in either 50 p.p.m. ALC or 200 p.p.m. tetracycline in sea water for 24 h in the dark to stain the shell margin. The stains did not cause mortality, and marks were restricted to the growing edge of the shells. However, since tetracycline resulted in a yellow fluorescent mark under ultraviolet radiation, which was not easily distinguishable from the shell color, ALC with highly defined red fluorescence was used in the present study.
Juvenile Manila clams smaller than 10 mm SL were collected from Lake Hamana (34°44′N, 137°34′E) in October 2005. The juvenile clams were brought back to the laboratory and marked with 50 p.p.m. ALC for 24 h. The marked juvenile clams were then stored in enclosures placed in intertidal and subtidal (10 cm depth at the lowest sea level at spring tide) environments in a tidal flat in Lake Hamana. The enclosures were made of 5-mm mesh plastic sheets rolled into 15-cm diameter cylinders, and were buried to the substratum with the top end exposed 10 cm above the ground. In each enclosure, 30 juveniles were stored.
The juvenile clams were successively recovered from the two stations 2, 5, 9 and 19 days after the beginning of the experiment on 15 October 2005. The shells were embedded in epoxy resin, cut and polished along the AMG, and the location of the ALC growth mark was checked under a fluorescence microscope. Acetate peels of the shells were made, and the number of increments in the outer shell layer from time of staining to recovery was counted to determine the formation periodicity by regression analyses. As in the case of adult Manila clam,13,15 thick and defined increments continuing to the middle shell layer were counted. The slope of the regression line was statistically tested by a Student's t-test.
Growth rate of wild juveniles
In order to confirm the modified acetate replica method, juvenile Manila clams were collected from the intertidal zone of the Seaside Park of Yokohama on 7 July 2005. The Seaside Park of Yokohama is an artificial tidal flat reclaimed in 1981, and it has high productivity of Manila clam despite recent low productivity in the other areas in Tokyo Bay.
For randomly selected juveniles (n = 28), SL was measured with a digital caliper to the nearest 0.01 mm, and AMG was measured on a computer screen connected to a light microscope to obtain the conversion formula to convert AMG to SL by regression analysis. Juveniles with SL of approximately 12 mm were randomly selected (n = 5), and shell acetate peels were analyzed for the growth trajectory ascending from the sampling date to as many days as possible by backcalculation using the AMG–SL conversion formula. Although the growth trajectory of AMG may not lie on a flat plane in the strict sense because of the growth pattern of the shells, possible measurement errors associated with this are ignored in this procedure. Growth increments in the outer shell layer were measured on a computer screen using a digital caliper to the nearest 0.01 mm.
Daily growth rate, daily percent growth rate (i.e. daily increment divided by the shell length in the prior day) and growth trajectory were analyzed. The pattern of daily increment formation (i.e. two-weekly rhythm associated with tidal activities) was also analyzed by Fast Fourier Transform (FFT) using the Microsoft Excel tool function (Microsoft, Redmond, WA, USA). Fourier values (imaginary numbers) were transformed into absolute values (F), and the Fourier power density was calculated as F2 to analyze the growth fluctuation pattern.
Acetate replica preparation
The section surface at the AMG of each juvenile Manila clam shell was polished with wrapping film sheets to remove scratches made by the preceding grinding by 2000-grade grit waterproof abrasive paper. Although polishing with 10 000 and 15 000-grade grits resulted in smoother surface, 8000 grade grit seemed sufficient for the light microscope observation.
As opposed to the conventional etching procedure for adult Manila clam shells (i.e. 3–5 min in 0.1 N HCl13), the combination of 0.05 N HCl and etching time between 10 and 30 s gave the best results for juvenile shells. Although ultrasonic cleaning seemed to successfully remove CO2 bubbles from the etching surface, it sometimes caused cracking of the shells.
Shells as small as approximately 2 mm SL could be treated with this procedure without many difficulties; however, it was difficult to expose the AMG of smaller shells even with the aid of a dissecting microscope, because of limited transparency of the epoxy resin. Growth increments were not always satisfactorily distinctive to assure constant counting results, and interpretations of the growth increments remained subjective.
Growth increment formation periodicity
Few juvenile clams were recovered dead from the field enclosure experiments; however, some were lost perhaps through the open top ends of the enclosures. Every recovered and examined juvenile clam had a clear ALC mark in the shell often accompanied by a disturbance band, indicating that the clam growth was halted by the staining procedure (Fig. 2). However, the staining did not cause sequelae, and the shell resumed its growth on the stained surface after the clam was placed in the enclosures.
Rearing day and the number of increments formed during the experiments were in significantly positive linear relationships both in intertidal (r2 = 0.96, n = 18, P < 0.001) and shallow subtidal (r2 = 0.91, n = 14, P < 0.001) stations; however, the data clusters were more precise in the subtidal (standard error, SE = 0.06) than in the intertidal (SE = 0.1) station (Fig. 3). The slope of increment count per day was not significantly different from 2 in both stations (Student's t-test, P = 0.077 and P = 0.81 for intertidal and subtidal, respectively), indicating that two increments are formed in one day on average.
Growth rate of wild juveniles
An allometric relationship was found between the AMG length (Mg) and shell length (SL) of the juvenile clams: SL = 1.18 × Mg0.96 (r2 = 0.98, n = 28, P < 0.001). This formula was used to convert Mg to SL in shell growth backcalculations.
The number of countable growth increments ranged from 141 to 161 among the five juvenile clams (9.6–11.8 mm SL) collected from the Seaside Park of Yokohama in July 2005. Limited resolution and erosion of the outer shell layer that occurred in the earlier growth period hindered further observation. The daily growth rate (i.e. sum of two consecutive increments converted from Mg to SL) was highly variable, ranging from 28.7 to 315.3 μm/day (five juveniles, n = 332), and the growth patterns of the individuals did not synchronize to one another (Fig. 4a). The mean daily shell growth rate of the five juveniles was 119.9 ± 44.6 μm/day (standard deviation, SD, 80 days), 120.9 ± 45.5 μm/day (59 days), 142.1 ± 58.8 μm/day (60 days), 130.9 ± 60.3 μm/day (67 days), and 124.9 ± 52.5 μm/day (66 days). The daily percent growth rate decreased with the increase in the shell size (Fig. 5a), and backcalculated growth trajectories based on the daily growth analysis were linear within the ranges of the shell size and time in the present study (Fig. 5b). The smallest and the largest minimum backcalculated shell lengths were 0.78 and 2.41 mm, respectively.
Although daily growth rates of the juvenile clams changed sinuously for several days' duration (Fig. 4a), frequency analysis of daily growth change by the FFT did not show any clear patterns (Fig. 4b). Frequency peaks were found in 6.3, 7, 7.9, 9, 10.5, 12.6 and 15.8 days in three individuals that had enough data for the FFT analysis (i.e. power function of 2, 64 days in this study).
A scientific basis for resource management based on biological and ecologic information is necessary to effectively stabilize fishery production of the Manila clam, the reasons for whose population decline have not been well understood in Japan. Although recruitment of marine organisms with high fecundity is said to be largely determined by mortality and growth rate in early stages of life history,16 little is known about the recruitment process of the Manila clam. The present study modified the shell acetate peel technique in order to analyze the growth of juvenile Manila clam, confirmed shell growth increment formation periodicity, and tentatively analyzed growth patterns of wild juveniles.
The acetate peel technique has been used for shell growth analysis of adult Manila clam.12,13,15 Treatment of fragile juvenile shells required more delicate grinding and polishing compared to the adult shells. Embedding the entire shell in epoxy resin enabled cutting and grinding without damaging the observation surface, and polishing the ground surface with an 8000-grade grit wrapping film sheet resulted in good resolution of juvenile growth increments. Etching conditions also differed from those used for the adult shells: 10–30 s etching with a 0.05 N HCl solution gave good results, as compared to 3–5 min with 0.1 N HCl for adult shells.13 A phase-contrast microscope gave enhanced resolution, making the increment identification easier; however, the acetate peels can also be observed with a regular light microscope. Use of ultrasonic cleaning during etching is not recommended for it may cause cracking of the shells. One could gently shake the epoxy blocks held by forceps or attached to a glass slide to remove CO2 bubbles formed during etching.
Daily and subdaily growth increments are considered to occur in the outer shell layer of the adult Manila clam, and the thick daily increments are distinguished from fine subdaily increments by distinctness and continuity to the middle shell layer.13,15 In the present study, however, distinct increments continuing to the middle shell layer were formed on every two days along with several indistinct increments in the field enclosure experiments. This correspondence between days elapsed and the number of distinct increments formed agrees with the results of a rearing experiment conducted by Richardson,12 in which he found that the Manila clam deposits clearly defined increments with an almost exact coincidence with the number of emersions under simulated semidiurnal conditions of emersion and immersion. Bivalve shells are formed by calcium carbonate deposition to the organic matrix (e.g. chonchiorin) secreted from the outer mantle epithelium.17 The Manila clam is indicated to have a semidiurnal endogenous metabolic rhythm by the oxygen consumption pattern.18 Thus, it seems reasonable to conclude that the Manila clam forms distinctive increments every two days in Japan's coastal areas with a mixed semidiurnal tidal pattern.
In the present study, the shell growth pattern in intertidal environments did not differ from that in shallow subtidal environments; therefore, shell deposition periodicity is not exclusively determined by emersion and immersion but perhaps by metabolic rhythm as a result of collective effects of various environmental factors associated with tidal movements. However, not only metabolic rhythm but also shell valve movements influence the shell growth.19 Indistinct shell increments that are not connected to the middle shell layer may be created by multiple valve movements within a day. Due to the presence of the indistinct increments, interpretation of the growth increments remains subjective, demanding experienced analytical skills. Upon use of the acetate peel method, one should observe and confirm the increment pattern with shells of known ages before conducting analysis of wild specimens.
The mean daily growth rate of the Manila clam estimated by cohort analyses is reported to be 86 μm/day (from 2.5 to 18 mm in 6 months) and 150 um/day (from 5 to 25 mm in 4 months) in Tokyo Bay (calculated from Murata20 and Nakamura et al.,21 respectively), and 94 μm/day (from 10 to 27 mm in 6 months) in Ariake Bay (calculated from Ikematsu7). Mean daily growth rate estimated from acetate peel analysis of adult Manila clam shells ranged 20–100 μm/day, depending on the ground height (i.e. faster growth in deeper zones) in the Banzu tidal flat in Tokyo Bay.13 The mean daily shell growth rates of juvenile clams collected from the Seaside Park of Yokohama in the present study ranged 120–142 μm/day, which falls within a reasonable range as compared with previous studies, supporting the validity of the modified method.
Shell growth increments of the cockle Anadara granosa have a tidal periodicity of formation,22 and the mussel Mytilus edulis has two-weekly fluctuations in shell growth associated with the spring–neap lunar cycle.23 This two-weekly pattern can be used for more approximate and less time consuming estimation of growth trajectory as compared with subdaily increments. However, the daily growth rate of the juvenile Manila clam showed a large fluctuation, and in the present study the FFT analysis did not detect a distinct growth pattern; therefore, two-weekly growth analysis may not be able to be performed on the Manila clam.
Although the modified method was not able to trace the shell growth back to the settlement size (i.e. ∼ 0.2 mm20) because of erosion of the outer shell surface and limited resolution, backcalculation up to 0.78 mm was possible. Despite its limitations, the use of shell internal growth increments has been shown to provide a reliable method in which juvenile Manila clam growth rate can be individually determined. Growth trajectories derived by this method can be used not only to study spatiotemporal factors determining the clam growth and the relationship between growth and death, but also to accurately identify cohorts by testing contemporaneity of individuals in a shell size frequency distribution. Reliable estimation of recruitment level based on cohort analysis is expected to help improve stock management of the Manila clam.
We are grateful to H Koike at Kyushu University for her technical advice. We also thank staff at the National Research Institute of Fisheries Science for laboratory assistance.