Population Characteristics and Sulfide Oxidizing Metabolism of the Bivalve Solemya velum

The physiological and biochemical studies of some animals from sulfide-rich habitats, such as deep sea hydrothermal vents, sewage outfall areas, eelgrass beds and mangrove swamps, have shown that sulfide-oxidizing chemoautotrophic bacteria exist endosymbiotically with the animals and provide a major energy source to their host. One of the remaining questions is the degree to which the symbiont supplies the host's need for reduced carbon and energy metabolism. Solemya velum, a common Atlantic clam from Nova Scotia to Florida, has chemoautotrophic symbionts in the gill tissue (Cavanaugh, 1983). The host animal has a reduced digestive system and lacks some digestive enzymes. The work reported in this thesis was intended to provide information on the population density and growth rate of the species in Ninigret Pond, Rhode Island, and on the energy input, energy output and energy consumption of the animal under the conditions of laboratory measurement. An energy budget based on a number of assumptions was proposed from both field and experimental data. The growth parameters L (length at infinity), K (the rate at ()() which the animal approaches L ), c (the intensity of the growth 00 oscillations) and WP (winter point) of the seasonal von Bertalanffy growth equation were estimated from the length-frequency of Solemya velum using an objective, computer-aided method (Electronic LEngth Frequency ANalysis, ELEFAN). The results were L = 19.6 mm, 00 K = 0.61, c = 1.0 and WP = 1.1 on a yearly basis. A value of total mortality Z = 2.63 was estimated for ii the adults from a length-conver t ed catch curve; therefore, the annual survival rate was 7.2% for the population. The highest density of the population 2 was 289 clams/m , and the average density was 120 2 clams/m . The other population characteristics, such as length vs. width, · length vs. dry weight, and wet weight vs. dry weight were also described. The enzymatic activities of sulfide oxidation were assayed in the gill tissues of Solemva velum. The maximal activities (V ) max were 4.7 umoles (s ubstrate converted to product) per gram fresh gill per min for thiosulfate transferase, 2.2 umoles/g/min for adenylylsulfate reductase, and 5.0 umoles/g/min for sulfate adenylyl transferase. The presence of sulfide stimulated and enhance d the respiratory rate of the animal. The animal respiration ( 65.2 ± 11.2 ul o 2 /g total wet weight (shell included)/hr) accounted 56.8% of the total oxygen consumption (115.4 ± 18.6 ul while chemoautotrophic symbionts consumed the rest 43.2% (50.2 ul An energy budget of Solemya velum, a chemoautotrophic symbiotic association, has been constructed based the growth rate, oxygen consumption and enzymatic activities of the species. The energy derived from sulfide oxidation may account 29% of the total energy input. The other sources of energy, like feeding and dissolved organic uptake, may provide the rest, 71%. The energy loss due to respiration and excretion may account 85% of the input energy, while only 15% may have been used for body growth. Multiple modes of nutrition are suggested for this protobranch bivalve. An energy flow model is proposed for the species, and may be applicable to some deep-sea hydrothermal vent organisms. iii

Florida, has chemoautotrophic symbionts in the gill tissue (Cavanaugh, 1983). The host animal has a reduced digestive system and lacks some digestive enzymes. The work reported in this thesis was intended to provide information on the population density and a length-conver t ed catch curve; therefore, the annual survival rate was 7.2% for the population. The highest density of the population 2 was 289 clams/m , and the average density was 120 2 clams/m . The other population characteristics, such as length vs. width, · length vs. dry weight, and wet weight vs. dry weight were also described.
The enzymatic activities of sulfide oxidation were assayed in the gill tissues of Solemva velum. sulfide-oxidizing enzymatic activities, will be submitted and to Biological Bulletin. The third one, on the bioenergetics of the species, will be submitted to Science.
The thesis as a whole is outlined t o unders t and the energetics in a chemoautotrophic symbiotic organism. Two limitations were encountered. First, no information on the proposed energy substrate (i.e. reduced ~ulfur) in the environment was obtained because of technical difficulty. Thus, the availability of the energy derived from reduced sulfur in that environment is not known. Second, this study provided a good picture of natural population characteristics and nutritional pathways, but in situ metabolism and reproduction are not known. I felt that the present results could stand by themselves with the questions remaining to be answered, serving as an incentive for further study .
vi Thesis Abstract.        . This genus of animals has a variety of unusual biological features. The species which have been investigated are all capable of rapid swimming and burrowing (Drew, 1900;. As observed originally by Pelseneer in 1891, the gut of Solemya is primitive and reduced or absent (Young, 1939;. In the gill tissues of a common Atlantic species Solemya velum, Cavanaugh (1983) has indicated the existance of intracellular bacteria, which are capable of utilizing reduced sulfur energy and fixing carbon.
Even though the biology and nutrition of Solemya have attracted attention for more than one hundred years, none of the studies have been directed to the population characteristics of the bivalve s. The present paper is intended to provide preliminary information on variation in such population characteristics as density, size-frequency, growth rate and biomass in the Ninigret Pond of Charlestown, Rhode Island.  (September 1984to July 1985. The core was 20.5 cm in diameter and 22 3 cm deep. The transects (one for each month) started from the shore and extended perpendicular l y to a water depth of approximately 1.5 meters, which was about 30 met ers from the shore. The whole sample area extended about 50 meters along the shoreline. The animals seemed to be distributed in a patchy pattern; almost no animals were found in the surrounding area 100 or 200 meters away from the sample site in the pond. Transects parallel with the shore were also taken. The cores were rinsed through a 5 mm sieve in field.
All individuals (>5 mm dia.) in each core were brought back to laboratory alive. They were counted, and measured to the nearest 0.01 mm in shell length and width with calipers. A total of 20 to 100 animals from each month's sample covering the full size range were weighed, dissected and freeze-dried for 48 hours. The dry soft-body and shell weights were measur e d to the nearest 0.01 mg on an electrical automatic balance.
The computer programs ELEFAN (~lectronic LEngth frequency ANalysis) developed by Pauly and David (1981) was used to analyse the length data to fit the best growth curve of the population. The original programs were written in Radio Shack BASIC II, and have been converted to BASICA in IBM personal computers. The von Bertalanffy growth function (VBGF) was used to describe the growth.
where Lt = length at relative age t, (1) L = length at infinity, 00 K describes the rate at which the animal approaches L 00 The parameters L and K of the VBGF were estimated on the basis of the length 00 frequency in Table 2. Because of the lack of information about age structure, the hypothetical age t 0 is set to zero, i.e. t-t 0 =t. The main features of the method (Pauly and Calumpong, 1984) include (1) It does not require other information besides length-frequency data, such as age classes or number of recruitment events per year; (2) it is based on the von Bertalanffy growth function which has been applied to a large number of species (including molluscs), and it uses the interpolative power of the equation to bridge gaps in the size-frequency; (3) the method has been applied to limited data sets among a number of species, e.g. seven-month length frequencies of a gastropod Dolabella auricularia (Pauly and Calumpong, 1984).
Polar and temperate . organisms generally display seasonal growth patterns. There is a routine in ELEFAN 1 which generates seasonally oscillating growth curves with two additional parameters. One is called Winter Point (WP), which corresponds to the time of the year when growth is slowest. The other is a dimensionless constant C, which expresses the intensity of the growth oscillations (Pauly and David, 1981). · Thus, the seasonal version of von Bertalanffy equation has the form where t sets the start of a sinusoid growth oscillation with regard to s t=O; the Winter Point is given by D describes the effect of animal size on growth, ranging from 0.3, such as in large tuna, to 1. O, such as in guppies . D was simply set to 1.0 in Solemya velum because of its small size. The value 6 of C can range from O in tropical fishes to 1 in temperate fishes. C=l is used as trial value in this case. Because the seasonal temperature fluctuation is more than 10 °c in Rhode Island (Figure 2), WP is set at 0.6 as a trial value as suggested by Pauly et al. (1983).
The computing procedure traces through a series Of length-frequencies sequentially arranged in time with a multitude of growth curves , and then selects a single curve which, by passing through a maximum of peaks, would "explain" these peaks. Six major steps are involved in the procedure (Pauly and David, 1981).
(1) Reconstruct the length-frequency samples. The procedure consists of calculating running average frequencies (over 5 length classes), dividing each length-frequency value by the corresponding running average frequency, then substracting 1 from the quotient. The positive values are considered as peaks, and the negative ones as troughs. (2) Calculate the maximum sum of peaks available in the length-frequency samples. This maximum "available sum of peaks" (ASP) is accumulated into a single growth curve.
(3) Analyse the length-frequency samples for a trial value of and K. A series of growth curves started from the base of each of The "slow" mode is used to search the system optimum (i.e. the best combination of parameters) by both switching starting points and modifying growth parameters. Before using the "slow" mode for searching the main growth curves, the "fast" mode is used to allocate a possible K range given a fixed set of L , WP, c, and starting point. The "fast" 00 mode iterates the growth parameters for only one starting Roint (the best one}, whil~ the "slow" mode would iterate both trial pa rameters and a starting point. When K is l e ss than 0.1, the ELEFAN faile d to identify the best combination of growth parameter values. The values of K less than 0.1 appear to be biologically impossible, so they were discarded.
K=0.6 was fi nally cho s en as a trial value. The trial L was set at 19.5 00 mm.
The ELEFAN II program is a complement to ELEFAN I. It estimates the total mortality (Pauly et al., 1982). The total mortality (Z) of the where L 1 and L 2 correspond to the lower and upper size limits of a class, respectively.
The length and weight were converted to logarithms and a linear regression was calculated. The regressions between length and width, wet weight and dry weight, length and dry shell weight were made for better defining the population. The resulting equation relating shell length to dry tissue weight were used with size frequency and density data to estimate monthly biomass.

At
where the unit of Lt is mm, t is the age in years after animal has spawned. The length was plotted against the age in Figure 4.
The estimated total mortality (Z) is 2.63. Z is defined as the coefficient of the instantaneous total mortality, which describes the rate at which the numbers in the population are decreasing (Equation (4)). This z value suggested that the annual survival rate (i.e. -Zt (Nt/N 0 ) = e , t=l yr.) of Solemya velum was 7.2% in Ninigret Pond. Figure 5 shows the relationship between shell length and dry tissue 2 weight on a log/log basis. The equation is a straight line (r =0.88): The biomass for five months available is shown in Table 3. The 2 highest value recorded was 17.3 g total we t wt/m ; the average biomass was approximately 6.7 g/m 2 . The dry shell weight (SW, mg) and shell length (L, mm) are also fit into a linear regression on log/log basis Note: (1). The densities were sampled for five months by the same coring device (20.5 cm diameter, 22 cm depth).
(2). The cores were taken along transects (one for each month) either perpendicular to or para l lel with the shoreline.
(3). The s.d. represents a standard deviation of each density sample. Note: 1) . The frequency is represented by the number of animals in each sample .
2) . --means that data was not available.  Note: the biomass was estimated from the densities in Table 1.

Seagrass beds
Seagrass beds are among the most productive biological systems known, typically producing between 500 to 1000 2 gC/m /year (Fenchel, 1977). Due to the high accumulation of organics and the lack of oxygen at depth in the sediment, the hydrogen and metal sulfides become concentrated. This typical feature of seagrass beds has some impact on the infaunal populations. Fenchel and Riedle (1970) stated that the sulfide biome is the only marine environment without crustaceans, as found in Ninigret Pond. Kikuchi and Peres (1977) reported that an impoverished infauna existed with Zostera marina (biomass of Zostera, 3500 g/m 2 wet wt), in which polychaetes (Spisula sachalinensis, biomass 88 g wet wt/m 2 ) tended to replace pelecypods, on the shallow coast of the Japan Sea. The average biomass (6.7 g total wet wt/m ) of Solemya velum may not seem important in terms of absolute biomass, but its average density (120/m 2 ) is relatively high. Because of its unique nutrition based on sulfide oxidation (Cavanaugh, 1983;) and the active locomotion behavior (Morse, 1931;Stanley, 1971), the ecological role of the animal in system energetics may be of special interest.

22
Growth equation The van Bertalanffy equation has been widely used in fisheries (Gulland, 1983). Considering the physiological _basis of the equation, it should be generally applicable to the animal kingdom, even though it may not fully describe a particular growth relation. The determination of growth through length-frequency analysis was based on the assumption that there was little year to year variation in growing conditions. The growth curves obtained estimate 'average' growth and in this sense they represent an integration of growth rates over slightly varying conditions. Pauly and Calumpong (1984) have obtained useful information on the life history of Aplysiid gastropods from field data by utilizing equations pertaining to fish. Thiesen (1973) suggested that the van Bertalanffy equation may only be valid for Mytilus (mussels) about one-third of their maximum size, and for small mussels the Gomperts K describes the rate at which the animal approches this limiting size.
The length of Solemya velum was given from 1/2 to 1 inch (12.5 mm to 25 23 mm) in a number of seashell books (Abbott, 1954 Generally, an animal of small size grows relatively fast. The hypothetical age t 0 can not be derived from the field data because of the lack of information on size of animals less than 5 mm. The seasonal growth oscillation was shown in the growth curve, which was generated from the growth equation (8)

Mortality
The high total mortality (Z=2.63) may be attributed to density effects, predation or environmental effects. The value of z is also an estimate of natural mortality, since there is no commercial fishing on these small clams. The annual survival rate was estimated to be 7.2%.
This survival rate applies to all age classes together; it may be lower for early age class individuals and higher for later age class individuals.

Chemoautotrophics
The sulfide-based chemoautotrophics is believed to be the major energy source for Solemya . This may provide the species a survival advantage over other bivalves in a sulfide-rich environment. Does this nutrition also change the growth mode of the animal? It would be of interest to conduct a comparative study of the growth pattern among comparable bivalves.
In summary, some of the methods developed from fish population dynamics have been applied to invertebrates, for example, the gastropod Dolabella auricularia (Pauly and Calumpong, 1984) (Cavanaugh et al., 1981;Felbeck, 1981Felbeck et al., 1981Felbeck and Somero, 1982;Cavanaugh, 1983). Schematically, this type of metabolism consists of the following pathways: sulfide oxidation, carbon dioxide fixation, nitrate reduction, and organic translocation and oxidation . Three diagnostic enzymes involved in sulfide oxidation are characteristic in many marine symbiotic animals from sulfide-rich habitats Fisher and Hand, 1984).
All were originally reported from sulfur-oxidizing bacteria (Kelly, 1982  . There are two kinds of factors which modify metabolism. One is related to the physiological or genetic constitution of the animal; the other is related to an ever changing environment (Hoar, 1983). One important factor of the former kind is body size; the sec,ond category includes the effect of substrate level, in this case, sulfide.
A shallow water bivalve, Solemya velum, was chosen to study the effect of sulfide on metabolism, and the activities of the sulfide-oxidizing enzymes. The percentage of oxygen consumption due to chemoautotrophics will be estimated. By making a few physiologically reasonable assumptions, the capacity of the organism to utilize reduced sulfur as an energy source will be estimated. The potential flux of sulfur through the species will be calculated. A comparison of energy metabolism in various vent animals, Solemya and Mercenaria wa s based on the work of Hand and Felbeck (1983). Salinity was approximately 28-32 parts per thousand. Six to eight animals were used in each experiment~ The length of the animals ranged from 10 to 15 mm.
A Gilson differential respirometer was used to measure the oxygen uptake rate in the presence or absence of hydrogen sulfide. The respirometer is a constant pressure system. The change in volume due to respiration is compensated by measured movement of a plunger in the enclosed space (Gilson, 1963). Three experiments were conducted at 15 °c 0 and one at 20 c (See Appendix A for detailed experimental conditions).
Hydrogen sulfide was made from sodium sulfide and diluted in filtered sea water. Crystal form of sodium sulfide was weighed by analytical balance. The seawater was purged with nitrogen gas to get rid of oxygen before the addition of sulfide. The solution was added to the vessels from side-arms. The initial concentration of sulfide in vessels was brought to 0.1, 0.5 and 0.8 mM respectively for three experiments. The initial sulfide was not known in one experiment. The three treatments were animal plus sulfide, animal only and sulfide only. Each treatment had three or four replicates. Four separate experiments were conducted.

33
The possibilty that sulfide oxidizing bacteria may exist outside of the animal soft-body or on the shells was not tested, but Cavanaugh (1983) has shown that the abundant existance of intracellular symbionts, which were sulfur-based chemoautotrophics.
Biochemical pathway of sulfide oxidation and their catalytic enzymes are presented in Figure 1. Three enzymatic assays were performed using fresh gill tissue which was homogenized in the buffer of appropriate pH (described below) with a Wheaton ground glass homogenizer~ The concentration of the extract of fresh gill tissue was 50 mg/ml. The crude homogenates were centrifuged at 5000 g for 15 minutes at o 0 c and the supernant fractions were used for the assays (Fisher and Hand, 1984 pentaphosphate in a total volume of 3 mls. The substrate (APS) concentration varied from 0.05 to 0.5 mM.
The method described by  was used for the assay of

Respiration
The weight specific metabolic rates of Solemya velum varied from 0.031 to 0.311 ul o 2 /mg total wet wt/hr. The total wet ~eight of the animals, including shell, ranged from 105 to 278 mg. The log-log regression of the weight specific respiration rates and weights was a negative linear correlation. Following Sutcliff's notation (1984), 2 log R = 0.9j -0~38 log W (r =0.86) s where R = weight-specific metabolic rate, W = weight. s (1) Animals showed a higher respiratory rate in the presence of hydrogen sulfide. Appendix B indicates the calculating procedure and lists the results obtained from four experiments using the Gilson respirometer, in which twenty eight animals were examined. The highest res p iratory rate, 0 which was observed when animals were exposed to 0.8 mM H 2 s at 20 C, was 215.96 ± 51.4 ul o 2 /g/hr. Because many other factors like body size, season, nutritional condition and sex, etc, would affect the metabolic rates , the results from different experiments were not comparable in this study. The variation of the respiratory rates in the absence of H 2 s were high, especially in experiments 1 and 2. The reason for this high variation may be that the bivalves often remained closed for the entire measurement, a major difficulty in working with bivalves.
But for all the experiments, animals always maintained a high metabolic rate in the presence of sulfide. Therefore, they all showed a stimulation and enhancement of animal-mediated sulfide oxidation.
The comparisons were done within the same experiment. Experiment 4 will be used for the comparison because of low standard deviations among 36 the replicates. The low variation in this experiment may have been due to the fact that field and laboratory temperature were not as different as in the ot.her three experiment. Figure 2 shows the weight-specifi c cumulative oxygen consumptions (ul/g) of three experimental treatments, animal with sulfide, animal only and sulfide only, from the experiment 4 with the Gilson respirometer. The mean and standard deviation of each treatment were calculated from four replicates.
In Table 1, the variation among four replicates and the differences between the paired variables, .i.e. respiratory rates in the presence and absence of sulfide over the whole time period, were calculated; and a t-test was run to see whether there was a significant difference between the two treatments. A t-statistic and a probability value for the hypothesis that the mean differences was equal to zero were listed in Table 1. The confidence level of the t-test was set to 0.05, therefore the differences between two treatments were all significant.   (Fisher and Hand, 1984). The activity of the enzyme was observed in six trials. The v of 4. 73 umoles /g/min (substrate max converted to product) was estimated from the linear portion of the reciprocal plot based on three trials, and is piesented with less confidence than APS reductase and ATP sulfurylase.  Table 2.   .
Only diagnostic enzymes are indicated. The enzymatic pathways may be expr essed as following reactions:

Biological features related to metabolism
All species of the Protobranchia are specialized for burrowing into and moving through a soft substratum . Solemya is especially adapted for the life under the surface, and its shell is modified for its habitat, g.g. it has an elongated shell with smooth surface. As  pointed out, "this genus appears unique in its capacity to live for a great part of the time without direct contact with the water above the substratum in which it burrows." He discussed in detail the evolutionary adaptation of the feeding and respiratory mechanism before the symbiotic chemoautotrophic nutrition was discovered in the animal Cavanaugh, 1983).
The unique biological features include: (1) The ctenidia (gills) are so large as to occupy about half the mantle cavity. The enlarged ctenidia are responsible for respiration . The dark color of the gills was noted as due to hemoglobin by .
(2) The gut of Solemya is reduced, and appears more primitive than that of the other protobranchia, like Nucula and Yoldia . The genus was often found in substratum of high organic content, g.g. sewage outfalls and eelgrass beds (Stempell, 1899;. (3). The Solemyidae probably represents one of the first lines along which the early Lamellibranchia evolved. They are a very old group appearing in the Devonian, and of the six genera only Solemya persists.
It is surprising that any species of a group, which evolved so early, 45 and along such specialized lines, have survived.
Based on the e x perimental results, some of physiological and biochemical function s of Solemya velum will be discussed, and correlated with the above uniqu~ biolog i cal features. Symbiosis may be responsible for the unusual features in the genus.

Respiratory response to hydrogen sulfide
The relation be twee n ani ma l si ze and metabol i c rates is a complex allometric adaptation (Hoar, 1983). It is well known from the studies of Zeuthen (1953) and Hemmingsen (1960) that the slope of a common regression line relating log metabolism to body weight of protozoans through to the larger poikilotherm metazoans is approximately -0.25 .
This slope varies from animal to animal, for example, in Mercenaria sp., the slope was-0.34 (Loveland and Chu, 1969); and the slope of a freshwater limpet was -0.27 (Berg et al, 1962). The slope of Solemya was -0.38 (Eq.(l)). The negative slope shows that younger or smaller animals have a higher rate of oxygen consumption on a weight-specific basis than older or larger ones. The relationship may be explained by the differences in the rate of enzyme production with increased size, variation in the cell surface to volume relationship (as proposed by van Bertalanffy, 1957), and in larger or old animals a greater production of tissues that have low metabolic rates (Prosser, 1975) .
In another protobranch bivalve Nucula turgida, Wilson and Davis (1984) found the absolute rate of oxygeh consumption in the range of . 0 temperature between 5-35 c was greater in summer-conditioned animals than in wi nter-conditioned ones. The oxygen consumption of winter 46 conditioned animal (Nucula) at 5°c was about 67% of that at 15°c. The animals in this study may be considered as winter conditioned because they were sampled at field temperatures of 0-7 °c. Therefore, their respiration at situ temperature may be approximately 60% of the measurements at 15°c. Even so, the measured temperature impact on each experimental animal was not estimated.
The presence of hydrogen sulfide stimulated and enhanced the metabolic rates of Solemya velum (Table 1). Hydrogen sulfide, in general, is a toxin to many organisms. It blocks the electron transfer system in the respiratory chain. Normally, Mollusca would simply close their shells and slow down metabolism . The possibility that there may have been sulfide-oxidizing bacteria on the outside of the animal soft-body or shells has not been excluded in this particular research, but it does not seem likely that these were abundant enough to have influenced the results. The outsides of the shells of Solemya were quite smooth and clean, and were further washed in clean sea water before measurements were begun. Cavanaugh (1983) observed rod-shaped intracellular symbionts (with characteristics of prokaryotic bacteria) in the gill tissue of Solemya velum; and that the presence of sulfide enhanced the carbon dioxide fixation in the gill tissues by 10 times.
Neither mantle nor foot has shown the same effect. It seems reasonable to believe that the difference of the respiratory rates in the presence and absence of H 2 s is due to symbiotic chemoautotrophics in the gill tissues of the animal.
The respiratory rates were different among the four experiments.
Reasons could include differences in season, size, sex, etc. Experiment 47 0 3 was conducted at 20 c, in which the respiration both with or without sulfide were two or three times higher than the other three (at 15 °c).
In experiment 2, low respiration was noted for both treatments, with and without sulfide. The respiration with sulfide was lower than that without sulfide for the first 90 minutes, and then the respiration with sulfide exceeded the other treatment for the last 30 minutes. this single experiment does not give a conclusive evidence of sulfide metabolism, but it has suggested that sulfide stimulated the respiration of the animals.
The impacts of temperature and sulfide on metabolism were also reported in Solemya reidi (McMahon and Reid, 1984). The weight specific oxygen uptake rate of Solemya reidi was reported to be inhibited below 6°C and above 18 °c; and was normal at 1.0 mM hydrogen sulfide, elevated at 0.5 mM, and ni l above 2.5 mM.
A c omparison of sulfide-enhanced respiration was made within Experiment 4 ( Table 1). The statistical tests show significant differences at a 0.05 confidence level between treatments in this experiment (t-test in Table 1 (Kuenen and Beudeker, 1982). For comparison, if the energy coupling efficiency between sulfide oxidation and co 2 fixation were 100%, the oxidation of every mole of sulfide would fix 1.36 mole co 2 .
According to the overall equation, the oxidation of each mole of sulfide needs two moles of oxygen.
In average, gill tissue weighed about 25% of total wet weight (including shell) . Therefore, cavanaugh's result is equivalent to 1.1 umoles co 2 /g total wet wt/hr. It should be noted that Cavanaugh used a different method (the uptake of labled carbon dioxide) and the gill tissues for her estimations, while the whole live organisms were used in this research. Of course, the assumption of 100% energy coupling efficiency is not realistic. A 20% efficiency was used for the energy coupling in thiobacilli (chemoautotrophic free bacteria) by Kuenen and Beudeker (1982). Thus, the co 2 fixation rate would be 0.30 umoles/g total wet wt/hr at a 20% efficiency, which is about one order magnitude lower than Cavanaugh's estimate (1.1 umoles co 2 /g total wet wt/hr). This may be due to the fact that Cavanaugh conducted her experiment in vitro.

49
Enzymatic activities in sulfide oxidation ; The enzymatic activities of sulfide oxidation in Solemya velum were in the range of values reported by other workers (Table 3). Felbeck et al (1981) reported a surprisingly high activity of ATP sulfurylase (77.0 umole/g fresh tissue/min) in Solemya reidi. The results of this study were very close to those from another bivalve, Lucina floridana (Fisher and Hand, 1984; see Table 3). Generally, the deep-sea vent animals, such as Riftia pachyptila (Pogonophona) and Calyptogena pacifica (Mollusca), have higher enzymatic activities. The hydrothermal vents provide a high, constant supply of hydrogen sulfide (Edmond et al, 1982), which serves as a major, if not the only, source of energy for the living systems. In shallow water, the percentage of energy expenditure derived from reduced sulfur is not clear, but may not be so large. Multi-nutritional modes may exist in shallow water s pecies, such as Solemya velum, as Yonge (1939) argued almost fifty years ago but without evidence.
In the pathway of sulfide oxidation (Figure 1), the number of enzymes involved, other than three diagnostic enzymes, is not clear. It is not known whether there is any enzyme catalyzing the first step from A study of sulfide oxidation in seawater suggested that the initial oxidation was simply a chemical reaction forming thiosulfate (Almgren and Hagstrom, 1974). Biological oxidation was subsequently responsible for the further oxidation of thiosulfate to sulfate (Tuttle and Jannasch, 1973). Thus, the above calculation and comparison are not intended to provide a true picture of this energetics, but rather to explore one possibility in an effort to find the actual pathway.
Adenosine phosphosulfate . (APS) is a central intermediate in the pathway (Kelly, 1982). It may be reasonable to assume that APS reductase is a key enzyme which controls the rate of the entire reaction. Also, the activity of APS reductase is the lowest among three enzymes in this study. For APS reductase, V = 2.2 umoles/g gill/min = 0.029 mole H 2 s/animal/yr, max given that the average size of experimental animals has 0.025 g fresh gill tissue in the total wet weight of 0.1 g (including shell). The As Hand and Somero (1983) concluded: "In general, the enzymatic activities found in the tissue of the vent animals were qualitatively and quantitatively similar to those of phylogenetically related sh~llow-living marine species, suggesting that the types of energy pathways and the potential flux rates through these pathways were similar in both groups." The important energy pathways include glycolysis, the citric acid cycle, and the electron transport system.
Many studies show that the activities of enzymes of energy metabolism correlate well with rates of oxygen consumption, even in organisms having widely different metabolic capacities, such as deep sea animals (Childress and Somero, 1979;Siebenaller and Somero, 1982).
The activity of Cytochrome C oxidase, which was found in the foot of Solemya, indicates a high aerobic capacity (Hand and Somero, 1983).
Phosphofructokinase and pyruvate kinase are indicators of total glycolytic flux potential. The high activities of malate dehydrogenase may be indicative of high activities for the type of anaerobic scheme found in many invertebrates.
Based on the above discussion, we may conclude that the unique character of Solemya is the pathway of sulfide oxidation, which generates energy for organic synthesis, while its energy metabolism remains similar to other bivalves, such as Mercenaria (Hand and Somero, 1983). Almgren, T. and I. Hagstrom. 1974. The oxidation rate of sulfide in seawater. Water research 8: 395-400. Bertalanffy, L. van. 1957. Quantitative laws in metabolism and growth. . sulphide oxidation enzymes in animals from sulfide-rich habitats.

Introduction
Host-symbiont interactions in molluscs may confer nutritional advantages on the host . symbiotic coral reefs will serve as a framework in the discussion .
The major components of sulfide-based chemoautotrophic symbiotic systems consist o f 1) a sulfide energy conversion system which produces ATP and NADPH, 2) carbon and nitrogen reduction systems which produces organics, 3) an organic translocating system, and 4) an metabolism system which produces biomass .  (Kelly, 1982). Kelly (1982) has reviewed the most likely alternative mechanisms of energy generation from thiosulfate among the thiobacilli. The same scheme can be probably applied equally to sulfide oxidation (Kelly, 1982).
The study of sulfide oxidation in seawater suggests that the initial oxidation (i.g. from HS to is simply a chemical reaction forming thiosulfate (Almgren and Hagstrom, 1974). Biological oxidation is subsequently responsible for the further oxidation of thiosulfate to sulfate (Tuttle and Jannasch, 1973). The principal energetic pathway of Solemya is assumed similar to the pathway of thiobacilli as indicated by the presence of the same enzymes. The APS-dependent phosphorylation is considered to be a control mechanism in the sulfide energy-yielding pathway of Solemya. Thus, APS reductase is considered as a control enzyme for the who le pathway.
The enzymatic activity of APS reductase was 2.2 umoles/g fresh gill/min (i.e. v ) . It had the lowest activity among the max three enzymes measured. Therefore, the estimation of energy flow based on APS reductase is conservative. One average size animal weighs about 0.1 gram (wet weight including shell), of which 25% is the weight of fresh gill . Thus, on the yearly basis, the maximum activity of sulfide oxidation would be 0.029 mole/animal/year.
The growth yield, 1.g. grams dry mass fixed per mole of thiosulfate, is 6.7 in aerobic thiobacilli (Kelly, 1982). The overall mean carbon content is 47% of dry biomass averaged from five species of Thiomicrospira (Kelly, 1982). The co 2 fixed to give that yield would be 60 0.262 mole (Kelly, 1982). Assuming that Solemya had similar conversion factors, for one average size animal (O.l g total wet wt), the symbionts could oxidize 0.029 mole reduced sulfur per year, and fix (0.029 mol H 2 S/yr) x (0.262 co 2 /H 2 S) = 0.008 mole co 2 fixed/yr, and (0.008 mol C/yr) x (12) x (l/47%) = 0.2 g dry biomass/animal/yr, which represents a theoretical productivity based on enzymatic activity.
The actual growth may be only a fraction of this estimate (See following section). It should be noted the growth yield is based on experimental measurements in free bacteria, therefore the energy transformation efficiency has been taken into consideration, even though it may be different from Solemya.

Growth of Solemya
The simplfied growth equation of Solemya velum is ( 1 _ e-0.6lt) Lt = 19.6 (1) where Lt is the length of animal at age t, t is relative age in years . Based on the equation, the yearly growth increment of a second-year animal is L 2 -L 1 = 13.77 -8.86 = 4.91 mm/year. converting length to dry soft~body weight by the equation , log W = -2.143 + 3.003 log L , thus the yearly increment of dry soft-body weight will be 0.014 gram/year based on the growth model of Solemya.

Respiration of Solemva
The oxygen consumption, which was measured in Gilson respirometer, included metabolic and chemoautotrophic portions when sulfide was present  The average effect of temperature on respiration was considered.
One year was divided into a winter averaged temperature of 5 °c and a summer averaged temperature of 20 °c. Wilson and Davis (1984) measured the oxygen consumption of the protobranch bivalve Nucula turgida under both summer and winter conditions in relation to temperatures (5-35 °c).
For a winter condition e d animal, they found that the consumption at 5 °c was 67% of that at 15 °c; while in a summer conditioned animal, the consumption at 20 °c is 180% of that at 15 °c. The respiration of Solemya was measured at 15 °c. Therefore, an approximate temperature factor over the year could be (0.67+1.80)/2=1.24. Here, I simply use 1 for the temperature factor for a conservative estimation in the energy budget of Solemya velum.

Excretion of Solemya
The ammonia excretion of Solemya velum has been measured by . He found a range of excretion rate of 7.212 to 9.541 umol NH 3 /g total wet wt/day. The mean value is approximately 8.4 umol NH 3 /g/day.
According to , the C/N ratio of protein metabolism is C/N = 0.529/0.173 = 3.06.

Energy budget of Solemva
The bioenergetic balance equation of a symbiotic association may be expressed as follows:  . Here, KJ=lOOO Joules. Therefore, E growth ( 3) where (W 2 -w 1 ) is the yearly increment of dry soft-body weight in grams; the carbon content of the animal was assumed to . be 40% of dry soft-body mass; the carbon content in grams was converted to moles by carbon molecular weight, 12; He is the carbon molar enthalpy of combustion. To satisfy the energy conservation law of thermodynamics in Eq (2), oxygen uptake was converted to the enthalpy of combustion equivalent of the catabolic organic mass, with consistently derived oxygen enthalpic equivalents (H 0 , KJ/mol o 2 ) , where N 0 is oxygen consumption by the animal in moles. The average values of substrate specific oxygen and carbon enthalpic equivalents H 0 and He were used : HO = 480 KJ/mol 02 He = 542 KJ/mol e .  has examined the ratio of heat production to oxygen consumption QH/Q 0 (KJ/mol o 2 ) in four marine bivalves. The ratio for The energy input calculated from the oxygen consumption seems to be reasonable with the comparison to the other two terms, growth and loss.
Many workers have shown that the activities of enzymes of metabolism correlate well with rates of oxygen consumption, even in organisms having widely different metabolic capacities, ~.g. deep-sea animals (Childress and Somero, 1979;Siebenaller and Somero, 1982). It is reasonable to assume that the energy input based on oxygen consumption represents a physiological capacity. The calculation indicates the chemoautotrophy may have provided 29% of the host's need. The energy derived from sulfide may be equal to 41% of the other energy sourpes, such as feeding and dissolved organic uptake. A total of 85% of the energy input may have been used for respiration and excretion, while only 15% may have been used for growth.
A conceptual model of energy flow in this chemoautotrophic symbiotic system is presented in Figure 1. The above discussion on the energy budget of Solemya velum may not be always true, but it does indicate that multiple modes of nutritions may exist in the species. More experimental efforts and theoretical calculations are needed in order to reconstruct the detailed features of the model. It may be interesting to compare those efficiencies to the energy flow model in Figure 1.
The measurement of stable carbon isotope ratio ( 13 c; 12 c) indicates that the chemoautotrophic nutrition is more important than the uptake of dissolved organic matter in Solemya reidi . The co 2 fixation mechanism in Solemya reidi appears to involve an initial trapping of co 2 into 4-carbon intermediate . The principal forms of reduced carbon and nitrogen, that are translocated from the symbionts to the host, are unknown.

67
Two approaches can be . used in the future to study the energy transfer in Solemya. The whole animal can be incubated in seawater containing 14 . C-bicarbonate, then labeled organic compounds can be seperated and identified in the host to describe the nature of carbon transfer.
The other approach is to analyse the nitrogen budget in the animal, like the studies in the symbiotic association of coral reefs . The carbon translocation rate was extremely high (90%) in a coral symbiotic association , but the overall transfer efficiency from primary producer to primary consumer approximately equalled the rate in actively foraging herbivores .
When balanced growth of carbon and nitrogen was assumed, it was found that much of the organic material translocated by zooxanthellae was deficient in nitrogen . The coral animal had a need to supplement its phototrophic carbon diet with nitrogen rich material--presumably from zooplankton or dissolved organic nitrogen compounds. By regulating the nitrogen nutrition of zooxanthellae, the animal host may control the symbiont's growth (Falkows ki et al, 1984).
It is possible that the nutritional interaction between symbionts and the animal (Solemya) would also regulate the growth of both. recently Solemya may have evolved mechanisms to prevent the poisoning of aerobic respiration by HS (Hand and Somero, 1983). In Rifitia pachyptila, a vent animal, HS is bound by blood-borne sulfide binding (transfer) protein thus preventing HS poisoning (Arp and Childress, 1983;Powell and Somero, 1983). Sulfide binding proteins may function both in protection of respiration, and in sulfide transport to endosymbionts. High activities of sulfide binding proteins were found in the foot of Solemya reidi which can oxidize HS to less toxic, or non-toxic, sulfur metabolites (Hand and Somero, 1983). This evidence may explain how Solemya can live in the habitats of high sulfide concentration, which normally blocks the respiration of organisms.
The animal makes Y-type burrows in sulfide rich mud .
It excretes a great amount of mucus during burrowing. The animal stays at the fork of the burrow, where both sulfide and oxygen are accessible.
The gill ctenidia of Solemya are so large as to occupy half the mantle cavity. This would not only provide a spatial microhabitat for symbionts, but also increase oxygen exchange area.
The gut of Solemya is reduced, and appears more primitive than that of the other protobranchia Reid and Benard, 1980). The reduction (g.g. Solemya velum) or absence (g.g. reidi) of the gut may be due to the development of 70 Solemya symbiotic sulfide-based nutrition. The controversy on the nutritional modes of Solemya remains today. Yonge found "the most minute particles" in the gut (1939).  f o und Solemya reidi takes up the dissolved organic matter in water, although it is less important than sulfide energy source.  claimed, "Solemya velum is a seston phage-filtrator feeding on the organic matter of the water entering the burrow ... The only species with the completely reduced intestine, Solemya reidi began to feed on the molecular organic matter dissolved in water." Solemyidae (family) are a very old group appearing in the Devonian, and of the six genera only Solemya persists. The species of Solemya are widely distributed . They are found on the east and west coast of North America, in West Africa, the Mediterranean . , the Canaries, Australia and New Zealand . It is surprising that any species of a group which evolved so early along such specialized lines have survived (Yonge, 1936). The symbiotic chemoautotrophy might have provided this genus with a significant survival advantage.
In summary, Solemya velum may have multiple nutritional modes. In the course of evolution, the species developed the nutrition of symbiotic chemoautotrophy. Thus, the original feeding nutrition became less useful, but might not have been abandoned. The species may also directly take up dissolved organic matter from water. The survival of the genus may indicate that the multiple nutritional modes have given the genus survival advantage, when the other genera in Solemyacea family went extinct.