Photosymbiosis in Clinocardium Nuttalli: Implications for Tests of Photosymbiosis in Fossil Molluscs Author(S): Douglas S

Photosymbiosis in Clinocardium nuttalli: Implications for Tests of Photosymbiosis in Fossil Molluscs Author(s): Douglas S. Jones and David K. Jacobs Source: PALAIOS, Vol. 7, No. 1 (Feb., 1992), pp. 86-95 Published by: SEPM Society for Sedimentary Geology Stable URL: http://www.jstor.org/stable/3514798 Accessed: 23/02/2010 16:24

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=sepm.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

SEPM Society for Sedimentary Geology is collaborating with JSTOR to digitize, preserve and extend access to PALAIOS.

http://www.jstor.org 86 RESEARCHLETTERS I

environmental correlates such as shallow, clear water in nutrient-poor, low latitude, reefal settings. When Photosymbiosis in these characteristics appear in fossil contexts, cases of symbiotic associa- Clinocardium nuttalli: tions are frequently inferred. This for Tests of approach has proven only moderate- Implications ly satisfactory, however, and has of- Photosymbiosis in Fossil Molluscs ten led to ambiguity (see Jones et al., 1988). The principal objective of this paper is to evaluate the traditional criteria used to identify molluscan photosymbiotic hosts in the fossil record. This evaluation is accom- DOUGLAS S. JONES photosymbiotic relationship of C. plished through: 1) an examination Florida Museum of Natural History, nuttalli (and other photosymbiont- of photosymbiosis in the modern University of Florida, bearing bivalves) contradict ecolog- heart cockle (Clinocardium nuttal- Gainesville, FL 32611 ical, morphological,geochemical, and li); and 2) a re-analysis of the tradi- life history characteristics tradition- tional criteria in light of the Clino- DAVID K. JACOBS1 ally used to recognize such associa- cardium results as well as recent work Department of Geological Sciences, tions in the fossil record. While these involving other photosymbiotic mol- Virginia Polytechnic Institute and characteristics remain valid as sup- luscan species. State University, porting arguments for suspected The heart cockle, Clinocardium Blacksburg, VA 24061 cases of paleophotosymbiosis, pale- nuttalli (Conrad, 1837), is widely dis- ontologists are cautioned that nu- tributed in littoral settings from San merous photosymbiotic taxa do not Diego, California, north to the Nu- PALAIOS, 1992, V. 7, p. 86-95 display these characteristics and nivak, Pribiloff, and Commander Is- hence many cases of paleophoto- lands, and south in the eastern Pa- The endosymbiotic association be- symbiosis probably go undetected. cific to Japan (Fraser, 1931). Although tween the heart cockle, Clinocar- Paleontologists are encouraged to it is of limited commercial signifi- dium nuttalli, and an endosymbiotic develop better criteria for recogniz- cance, this shallow burrower is com- green alga (zoochlorella) was inves- ing photosymbiotic molluscan hosts. monly collected by recreational clam tigated in the field and laboratory. diggers from intertidal sandy mud Specimens were obtained from False flats and sand beaches throughout its Bay, San Juan Island, Washington. INTRODUCTION range. Clinocardium nuttalli is also Infaunal clams did not contain sym- one of a small but growing number bionts while slightly larger and older Identifying photosymbiotic rela- of molluscan species that has been semi-infaunal individuals harbored tionships in the fossil record is prob- reported to harbor endosymbiotic al- algae in siphonal and mantle tis- lematic because the symbionts are not gae (Hartman and Pratt, 1976); it is sues. Epifaunal clams were the larg- preserved, nor do they produce iden- this relationship that is of particular est and oldest (up to six years old) tifiable structures in the hard parts interest here. and contained the greatest concen- or skeleton of the host (Cowen, 1983). Algae and chloroplast symbionts tration of algae. The three mode-of- Because direct evidence of such re- are known from only two modern life groups apparently represent an lationships is lacking, indirect crite- molluscan classes, the bivalves and ontogenetic continuum with pro- ria (i.e., analogies with modern taxa) the gastropods. Three types of sym- gressive emergence from the sub- are traditionally used to argue for bionts are involved: 1) zoochlorellae strate related to the acquisition of suspected cases of photosymbiosis. (Chlorophyceae, Chlorocaccales); 2) algal symbionts after the clams' sec- These arguments have been applied zooxanthellae (Dinophyceae, Peridi- ond year of life. No symbiont-in- to diverse taxa, but especially to fos- niales); and 3) free chloroplasts (de- duced effects were observed on either sil molluscs. rived from several plant orders) shell growth rates or on oxygen and Selected ecological, morphological, (Muscatine and Greene, 1973). En- carbon isotopic composition of shell geochemical, and life history char- dosymbiotic chloroplasts occur in carbonate. Several aspects of the acteristics have been associated with several species of marine, hermaph- modern photosymbiotic molluscan roditic snails of the order Sacoglossa hosts. Among these are: 1) special- (Gastropoda,Opisthobranchia). More ized which enhance 1Present address: Department of Inverte- morphologies familiar are the endosymbiotic algae brates, American Museum of Natural His- light exposure; 2) rapid calcification (dinoflagellates) known as zooxan- tory, Central Park West at 79th St., New producing large skeletal sizes; 3) frac- thellae. These occur among a wide York, NY 10024. tionation of carbon isotopes; and 4) range of opisthobranchs, but are best

known for their sustained, intimate association with bivalves of the su- perfamily Cardiacea, particularly the family Tridacnidae (Muscatine and Greene, 1973). Associations involving zoochlorellae are generally less familiar than the other two groups and often in- volve algae which become parasitic (Muscatine and Greene, 1973). The term "zoochlorellae" has no real tax- onomic significance and has traditionally been used to refer to those green algae found as endosymbionts in freshwater invertebrates (Trench, 1979). For example, zoochlorellae are known from the freshwater bivalves Anodonta cygnea and Unio picto- rum where they are confined to the tissues that receive maximum illu- mination, such as the posterior mantle edge, siphons, and occasionally the FIGURE1 -Map of northwesternWashington with expanded view of San Juan Island(inset) foot. Zoochlorellae have also been indicating location of Friday Harbor Marine Laboratory and collection site of Clinocardium identified in marine molluscs. They nuttalliat False Bay. were reported as mantle endosymbionts in the giant scallop, Placopec- ten magellanicus (Naidu and South, ognize such relationships in the fossil subset of both mantle and siphonal 1970; Naidu, 1971), and more re- record are re-evaluated in the dis- tissues containing symbionts was se- cently as endosymbionts in the man- cussion that follows. lected for examination by transmis- tle, siphon, and occasionally the foot sion electron microscopy (TEM) to of the heart cockle, Clinocardium MATERIALS AND METHODS investigate the nature of the symbi- nuttalli (Hartman and Pratt, 1976). ont in more detail. These tissues were In general, these associations involv- Individuals of Clinocardium nut- pre-fixed in 3% buffered glutaral- ing zoochlorellae have not been in- talli were collected by hand from the dehyde at the FHML and were later vestigated as extensively as those in- lower intertidal zone at False Bay, dehydrated, sectioned, and stained volving other algal endosymbionts. San Juan Island, Washington (Fig. according to standard techniques. The intimacy of algal-molluscan 1). Specimens came from the fine sand The TEM analyses were performed symbiotic associations varies from and silt tidal flat environment, within on a JEOL 100C instrument at facultative to obligate and from tran- about 0.3 m of mean low water. Clams VPI&SU, Blacksburg, Virgina. To sient to sustained. In most cases the with and without algal symbionts further aid in the identification of the relationships are presumed to be mu- were collected. The former were rec- algal symbiont, tissues were stained tualistic with the host benefitting ognized in the field by the green color for the presence of starch using Lu- from symbiont-derived photosyn- of the siphonal region, or in the case gol's solution. thates released inside its tissues. The of epifaunal individuals, by the The soft tissue from each specimen symbiont benefits by living in a ho- greenish mantle tissue exposed by the was ultimately removed and the meostatically-controlled microenvi- gaping valves. Both epifaunal and valves were saved for analyses of shell ronment where its requirements for semi-infaunal clams harboring en- growth increments and 13C and 180 light, nutrients, inorganic carbon dosymbionts were recovered, as were content. The 17 sets of paired valves, (CO2), and trace elements are pro- infaunal specimens which uniformly five each from the epifaunal and semi- vided (Trench, 1979). However, the lacked symbionts. The latter were infaunal groups and seven from the actual fluxes of carbohydrate (14Cla- normally found slightly higher in the infaunal group, were washed and pre- belling) or nutrients have only been intertidal zone. Seventeen specimens pared for cross-sectioning. The right documented in a fraction of the cases were collected and transported to the valve of each pair was radially sec- of modern photosymbiosis. Without Friday Harbor Marine Laboratory tioned from the umbo to the ventral such measurements, a range of sym- (FHML) for further investigation. margin along the axis of greatest shell biotic relationships can be inferred. In the laboratory, all specimens height. The radial cross-sections ex- Mutualism is implied by adaptive be- were measured and labelled and the posed annual shell growth bands havior or morphologies enhancing distribution of symbionts was ob- which were used to estimate the age symbiosis. The criteria used to rec- served on a macroscopic level. A of each individual and to guide sam- 88 JONES & JACOBS I

ANNUALGROWTH BAND inner shell layers of the right valve of one specimen and isotopically analyzed according to the methods de- scribed above.

RESULTS Nature and Identity of the Symbiont TEM documents the existence of endosymbiotic algae in the siphonal and adjacent mantle tissues, as well as in portions of the foot immediately ISOTOPIC SAMPLE below the siphon of the host, Clino- FIGURE 2-Radial cross-section through right valve of Clinocardium nuttalli showing annual cardium nuttalli. Colonies are often growth bands and positions of samples drilled for stable isotopic analyses. merged into larger cell masses with tetrads of replicating cells present (Fig. 3). The lack of darkly stained pling for stable isotopic analysis (e.g., move organic contaminants. X-ray nuclear material and the vivid green Jones et al., 1983). diffraction analysis of powders before (rather than brownish) color of the One sectioned valve from each and after roasting confirmed that only symbionts when viewed under light mode-of-life group was selected for aragonite was present. The samples, microscopy rule out the possibility stable isotopic analysis (180/160, 13C/ analyzed according to standard tech- that this symbiont is a dinoflagellate. 12C)of shell carbonate. Isotopic anal- niques (Williams et al., 1977), were Additionally, eight or more thakloids yses were undertaken to help evalu- then reacted in vacuo with concen- are grouped in grana. Grana consist- ate the proposal that the shells of trated orthophosphoric acid at 70? C. ing of more than three thakloids are molluscs harboring photosymbionts An on-line, automated carbonate characteristic of green algae (Steve display unique carbon isotopic sig- preparation system facilitated the Herbert, pers. comm.). The electron natures (Cowen, 1983). The outer production and purification of the micrographs are similar to those pub- shell surface of each sectioned valve evolved CO2 gas. The isotopic com- lished by Hartman and Pratt (1976, was first ground down until fresh shell position of the CO2 was determined fig. 4), who concluded that the alga was exposed and all of the periostra- with a fully-automated, VG Isogas infecting C. nuttalli from Yaquina cum had been removed. A small den- PRISM Series I mass spectrometer. Bay, Oregon, was most probably a tal drill with bit size <0.5 mm was All values are reported in standard 6 green alga of the form genus Chlo- then used to remove aragonitic shell notation as the ratio (R) of 180/160 or rella, an identification compatible material for isotopic analysis. Eigh- 13C/12Cin the sample to that of the with the observations here (Lynn teen samples (plus two duplicates), PDB standard according to the equa- Margulis, pers. comm.). each weighing approximately 0.5 mg, tion: Green algae normally contain were obtained from each shell. The but with so- 6180 or = starch, staining Lugol's sampling pattern was identical for all 613C(%0) [(Rsple/Rst,drd) lution revealed none. A similar - 1] x 103 neg- shells (Fig. 2): 1) eight consecutive ative result was encountered by Nai- samples (spaced about 1 mm apart) Results were calibrated to PDB using du and South (1970) who stained for were drilled from the outer shell lay- daily analyses of NBS-20 and/or the presence of starch in the green er, beginning at the ventral margin NBS-19. Precisions for standard algal endosymbionts of the sea scal- and extending toward the umbo; 2) analyses are ?0.11%o for 6180 and lop, Placopecten magellanicus, from five consecutive samples were drilled ?0.06%oofor 613C. Newfoundland. Perhaps in such from a region of the outer shell layer Dr. Susan Kidwell (Dept. of Geo- symbiotic associations, metabolite near the umbo, within the first year physical Sciences, University of Chi- storage occurs in the host rather than of shell growth; and 3) five samples cago) supplied several specimens of the algae, explaining the lack of starch were ground from the interior surface the venerid, Protothaca staminea, storage. of the shell, within the pallial line and collected at about the same time and at inter- from the same mud flat as the spaced approximately equal spec- Field and Laboratory vals across the shell section. This imens of Clinocardium nuttali. Pro- Observations sampling strategy was designed to tothaca staminea does not harbor permit comparison of isotopic ratios endosymbiotic algae and was used as Infaunal specimens of Clinocar- from various parts of the shell. a control species for comparison of dium nuttalli, found in the upper Each powdered sample was roast- isotopic values. Ten samples were reaches of the intertidal zone, did not ed in vacuo for 1 hr at 350? C to re- drilled from the aragonitic outer and contain endosymbiotic algae. These PHOTOSYMBIOSISIN FOSSILMOLLUSCS 89

FIGURE 3-TEM photomicrographs of endosymbiotic green algae of the form genus Chlorella living both extracellularly and intracellularly in the tissues of Clinocardium nuttalli. A) Algal cells living extracellularly within the haemocoel. Scale bar equals 0.8 Am. B) Intracellular algal cells and tetrads. Scale bar equals 1.5 Am. C) Single cell from mantle of C. nuttalli containing encapsulated algal cells. Scale bar equals 1.2 Mm.D) Single intracellular tetrad. Scale bar equals 0.5 Am. specimens were on average smaller in clumps in those tissues exposed to with their valves agape, exposing a than their semi-infaunal or epifaunal light, according to the mode of life of maximum amount of tissue to sun- counterparts (Fig. 4). Annual, inter- the individual in question. Semi-in- light. Accordingly, symbionts were nal shell banding (discussed below) faunal clams contained symbionts in widely distributed throughout the si- suggests that these infaunal individ- the siphonal region and in the mantle phon and adjacent mantle as well as uals were also younger, two years old immediately adjacent. Some of these in the foot and gill supports and with- or less. Semi-infaunal clams lived at individuals exhibited a few clusters in the mantle interior to the pallial or below mean low water and all con- of symbionts on the surface of the line. Much of the external shell sur- tained symbionts. They were notice- foot facing the exposed siphonal por- face of epifaunal clams was colonized ably larger than the infaunal clams tion of the commissure. Ulva, as well by Ulva, indicating a prolonged epi- and had a greater number of annual as filamentous green algae, had col- faunal existence. Infaunal clams not bands. Epifaunal clams, all contain- onized the posterior margin of semi- only lacked endosymbionts, their ing photosymbiotic algae, were still infaunal individuals, suggesting that shells were also devoid of macroscop- larger and older (Fig. 4). these clams rarely if ever retreat into ic algae and had a yellowish hue rath- On a macroscopic level, the green the sediment. er than the dark gray-brown of pho- algal endosymbionts were observed Epifaunal individuals were found tosymbiotic individuals. Epifaunal 90 JONES & JACOBS

clams did not display symptoms of ter within the valves) noted in soime in each of the three mode-of-life ill health (e.g., decline in condition heavily infected specimens of Placco- groups are shown in Figure 5. Though of visceral mass and adductor mus- pecten magellanicus (Naidu, 1971,,p. based on small sample sizes, the cle, loss of total body weight, exces- 148-151). Instead, specimens hlad curves are essentially indistinguish- sive slime development around af- firm tissues of normal size and col,or, able, suggesting that no appreciable fected tissues, inability to maintain were able to achieve tight valve c]lo- differences exist between the growth tight shell closure and retain free wa- sure, and were vigorous to the poiint rates of the three groups and that the of "leaping" out of shallow fingger semi-infaunal and epifaunal clams are bowls of relaxant by rapid extensi ion larger because they are, on average, Shell LengthGrouped According to Mode of Life of the foot. older. In fact, the specimens do not Radial cross-sectioning of t he represent distinct populations, but I zI I I valves revealed a macroscopic, altEer- rather (as discussed later) an onto- EPIFAUNAL nating pattern of growth bandi ng genetic sequence involving progres- {I FT-j H which was interpreted to form wiith sive emergence from the sediment SEMI-INFAUNAL an annual periodicity. The comlbi- that is linked to the acquisition of I E- - I nation of one thin, translucent ba]nd photosymbionts. INFAUNAL (dark colored in reflected light) a]nd I ,,1 , 1 1 , one wide, opaque band (whitish c(ol- Stable Isotopes 3 4 5 6 7 8 ored in reflected light) in the outter Length (cm) shell layer comprises a yearly cy(cle A scatterplot of the (180 versus 613C FIGURE4-Box andwhisker plots of shell of shell growth (Fig. 2). This patte rn ratios for all isotopic samples from lengthsfor infaunal (n = 7), semi-infaunal(n is well documented in a wide varie.ty each of the three mode-of-life groups = = 5), and epifaunal(n 5) specimensof of marine bivalves, including cock]les of Clinocardium nuttalli as well as Clinocardiumnuttalli. Each verticalbar rep- (e.g Is, for Protothaca staminea is shown in resentsthe median value for each group, the ., Jones, 1983 Lutz and Rhoa( box represents one quartileof the data on 1980). Figure 6. The extensive overlap of the eitherside of the median,and the whisker Representative growth curves f'or oxygen and carbon isotopic values for representsthe range. specimens of Clinocardium nutta ~lli all clams suggests that there are no appreciable differences between the isotopic records of C. nuttalli from 80 each mode-of-life group or between C. nuttalli and the control species, P. staminea. These generalizations were confirmed by statistical analyses comparing the means of the various distributions of isotopic ratios. 60 Shown below the scatterplot in E Figure 6 are the mean and 95 % confidence interval for the b13Cvalues of each group. The large overlap of the confidence intervals indicates no statistical difference between the means w 40 of these groups. The least amount of overlap is seen between infaunal Cli- nocardium and Protothaca; yet, a w~.1 t-test comparing the mean (13C val- I ues from these two groups revealed C,, 20 no differences at P < 0.05. Similarly, t-tests comparing the mean 13C(and 6180) values of specimens with endosymbionts (epifaunal plus semi-infaunal C. nuttalli) against the mean values of specimens without endosymbionts (infaunal C. nuttalli plus 0 2 4 6 8 P. staminea), revealed no statistical differences between the mean carbon AGE (years) (or oxygen) isotopic ratios (P < 0.05). FIGURE 5-Size (shell height) versus age relationships for Clinocardium nuttalli from False The smooth variation in oxygen Bay. Vertical bars indicate ?1 SD. Triangular symbols denote infaunal clams (n = 7), circ;les isotopic ratios of the eight consecu- denote semi-infaunal clams (n = 5), and squares denote epifaunal clams (n = 5). tive samples recovered from the mar- PHOTOSYMBIOSISIN FOSSIL MOLLUSCS 91

ginal region of each specimen of Cli- 1 nocardium is thought to reflect an annual periodicity in 6180 related to 0E Infaunal 0 0 seasonal shell growth; however, the [] 0 U eight samples are insufficient to de- Epifaunal 0 a tect complete cyclicity and more de- [] Semi-infaunal tailed sampling is needed to confirm A Protothaca this hypothesis. Samples drilled in the dark growth bands had the most en- 0- A 0? riched values for each 0 oxygen isotopic 0 X o 0 clam (+0.57, +0.70, +0.70), suggest- 0 ing that the dark bands formed during the coldest time of the year. This []a A has been observed in a va- 0 X El pattern 0 0 0 riety of marine bivalve species which 1- 0 have been in detail * o serially sampled CD - -1 0Q (e.g., Jones et al., 1983; Krantz et al., A 0 A 1984, Jones et and A 1987; al., 1989) o[] supports the interpretation of the 0 major growth bands as annual fea- 0 tures. 0 DISCUSSION

_- , I * s Photosymbiosis in /I I\ Clinocardium nuttalli -2 0 ' 1 \ The results of our investigation of \N photosymbiosis in Clinocardium N / 813C (?/oo) nuttalli from False Bay, Washing- N ton, confirm the earlier observations -1. -0.5 m/ by Hartman and Pratt (1976) based IL V on specimens from Yaquina Bay, Or- egon, and expand our understanding Infaunal I E I of the algal-molluscan relationship. The TEM and light microscopic in- Epifaunal [ o i vestigations support the identifica- Semi-lnfaunal I e I tion of the symbiont as a zoochlorella

Protothaca I &_ I (green alga), probably of the form ge- a...... s nus Chlorella. Furthermore, field ob- FIGURE6-Scatterplot of 6180 versus 613C from shells of infaunal (open squares), semi-infaunal servations and size-age relationships (dark squares), and epifaunal (circles) Clinocardium nuttalli. Triangular symbols indicate iso- confirm Hartman and Pratt's (1976) topic values from Protothaca staminea, which served as a control. Below scatterplot are the observation that the onset of pho- mean and 95% confidence intervals of the 613C values for each group. tosymbiosis in C. nuttalli occurs later in the ontogeny of this animal than in the more familiar symbiotic rela- suggest that these clams do not fre- to larger, semi-infaunal clams which tionships involving zooxanthellae and quently rebury and that they first ac- acquire symbionts, to large, epifau- molluscs such as Tridacna (Trench quire symbionts in the siphonal tis- nal individuals with extensive sym- et al., 1981). sue during the semi-infaunal state. biotic infections, is likely to represent Small, young, infaunal clams lack The symbionts later spread from the an ontogenetic progression. The on- endosymbionts. Larger (Fig. 4) and siphonal region throughout the man- set of photosymbiosis occurs after the older (Fig. 5) semi-infaunal clams tle and foot in older, epifaunal clams. second year of life in conjunction with contain endosymbionts only in those The more extensive algal overgrowth a gradual emergence from the sedi- tissues exposed to sunlight, particu- on the shells as well as the distribu- ment. This is consistent with the larly around the siphonal region. On tion of symbionts indicates these epi- findings of Hartman and Pratt (1976) the shell exterior, the pattern of en- faunal clams do not retreat into the who reported that no individuals less crusting algal overgrowth concen- sediment once they have made the than two years of age contained sym- trated on the posterior, emergent transition to the surface. bionts. Their laboratorystudies which portion of the shell. This pattern plus The pattern of small, infaunal exposed clams to fresh and cultured the distribution of endosymbionts clams without symbionts progressing zoochlorellae indicated that the young 92 JONES & JACOBS

clams (1-year-olds) phagocytized the went through a neotenous transfor- li have a number of implications re- algae and remained free of symbionts mation leading to the retention of a garding the criteria that are normally whereas 50-75% of four-year-olds large byssus and eventually to a sed- used to identify cases of molluscan became infected. entary, bysally attached mode of life photosymbiosis in the fossil record. It is difficult to demonstrate un- where symbiosis developed much As mentioned earlier, these criteria equivocally that the clams derive nu- earlier in ontogeny. fall into four basic categories: 1) mor- tritional benefit from their symbi- Finally, shell carbonate of Clino- phologies enhancing light exposure; onts without performing pulse chase cardium nuttalli was examined to test 2) calcification rates; 3) stable iso- experiments with 14C-labelled CO2. whether algal symbionts had any ob- topes; and 4) paleoenvironmental The ontogenetic emergence sequence servable effect on isotopic composi- considerations. A number of these in Clinocardium suggests that the tion. Many photosymbiotic host criteria have already been criticized clams emerge from the sediment to organisms such as corals and fora- in terms of their general applicability facilitate symbiont infestation of ad- minifera exhibit disequilibrium frac- (Jones et al., 1988) and are reconsid- ditional mantle tissue. Symbiotic in- tionation of 13C and 180 in their car- ered below in light of C. nuttalli and dividuals remained vigorous and bonate skeletons (e.g., Weber and other photosymbiotic species: healthy. There is no evidence to sug- Woodhead, 1970; Weber et al., 1976; gest that the clams were incapable of Erez, 1978; Williams et al., 1981; Morphologies Enhancing reburial. Reburial would eliminate Swart, 1983). Such occurrences led Light Exposure the photosymbionts if they were dis- Cowen (1983) to propose that similar advantageous. These observations are effects might be observed in photo- According to Cowen (1983), if a in agreement with Hartman and symbiotic molluscs. A subsequent in- fossil species maintained a charac- Pratt's (1976) assessment that there vestigation attributed lower 613C in teristic life position or possessed spe- were no detrimental effects on C. the bivalve Tridacna maxima com- cialized structures which could be nuttalli as a consequence of algal in- pared to a coeval gastropod to the shown to enhance tissue exposure to fection. These observations also sug- presence of photosymbiotic zooxan- light, then an hypothesis of algal gest that C. nuttalli acts to accom- thellae (Jones et al., 1986; Romanek symbiosis could be proposed. Famil- modate the symbionts and is likely et al., 1987); however, a later study iar examples among modern bivalves to maintain a mutually beneficial re- concluded that zooxanthellae in Tri- include the body orientation and lat- lationship with them. dacna do not impart an identifiable eral flanges of Tridacna as well as the In discussing evolutionary trends effect on the isotopic composition of living position and the "windows" and phylogenetic development among the shell carbonate (Romanek and through the shells of Corculum which the bivalve molluscs, Seilacher (1984, Grossman, 1989). promote light transmission (Vogel, 1985) argued that nutritional bene- As previously discussed, no sym- 1975; Watson and Signor, 1986). An fits of photosymbiosis represent a biont-induced fractionation effects unfortunate limitation of this crite- strong enough selective force to have were noted in the isotopic records of rion, however, is that the converse lured (evolutionarily) particular lin- Clinocardium nuttalli. Comparisons does not always hold; that is, pho- eages of bivalves back out of the pro- were made between infaunal, apo- tosymbiotic bivalves do not always tection of the sediment into an epi- symbiotic clams and semi-infaunal possess specialized morphologies or benthic mode of life. Our evidence and epifaunal clams with symbionts. life positions related to their photo- suggests that photosymbiosis in Cli- Intra-specimen comparisons be- symbiotic habit. Clinocardium nut- nocardium nuttalli develops early in tween the isotopic ratios of the ju- talli is a good example. If this species ontogeny and does not appear in all venile (aposymbiotic) stage and the were extinct and known only from members of the population, appar- mature (post two-year-old, symbi- fossil shells, there are no morpholog- ently being restricted to clams living ont-hosting) stage of these latter two ical features (other than evidence near or just below mean low water. groups also revealed no differences. such as epibionts, borings, shell pit- These observations suggest that pho- When comparisons between the ox- ting, etc., which suggests an epifaunal tosymbiosis is in statu nascendi in ygen and carbon records of all of the existence) that would prompt a pa- this particular taxon, with clams in C. nuttalli specimens were made with leontologist to propose C. nuttalli as the process of being lured from the the corresponding records from sym- a photosymbiotic host. The same can sediment. From an evolutionary per- biont-free Protothaca staminea from be said of the sea scallop, Placopec- spective, it is possible that C. nuttalli the same habitat, again, no differ- ten magellanicus, which has been re- is at an intermediate stage on an evo- ences were observed. ported to harbor endozoic algae (Nai- lutionary path that leads to photo- du and South, 1970; Naidu, 1971) or symbiosis. A comparable scenario Recognition of Photosymbiosis the unedo cockle, Fragum unedo, a could be envisioned in the evolution- in Fossil Molluscs: small, shallow burrower that spreads ary history of tridacnids, for exam- Criteria Re-Evaluated its zooxanthellae-laden mantle over ple, where a cardiid-like ancestor be- the sediment like a carpet (Kawaguti, came photosymbiotic in a manner The results of the study of pho- 1983). While the presence of special- similar to C. nuttalli. It subsequently tosymbiosis in Clinocardium nuttal- ized, light-enhancing structures or PHOTOSYMBIOSISIN FOSSIL MOLLUSCS 93

body orientations may suggest in- Stable Isotopes oligotrophic (reefal) environments stances of photosymbiosis in fossil and in fact, Hallock and Schlager molluscs, the number of modern, During photosynthesis, carbon and (1986) suggest that calcifying pho- symbiont-bearing taxa without ei- oxygen isotopes are fractionated. tosymbiotic organisms (especially ther attribute suggest that many cases Evidence of this fractionation, par- those constituting coral reefs) are un- of photosymbiosis in the fossil record ticularly with respect to carbon iso- likely to propagate extensively in the go unrecognized. topes, is often preserved in the car- absence of such conditions. These bonate skeletons of organisms hosting general environmental characteris- photosymbionts such as foraminifera tics do not always fit the settings in and corals (e.g., Erez, 1978). It has Calcification Rates which we find modern photosym- been suggested that putative cases of biotic molluscs, calling into question It is generally stated that hosts of photosymbiosis among fossil mol- the predictive utility of such paleoen- algal symbionts secrete carbonate luscs might be identified by search- vironmental criteria in identifying more rapidly than their aposymbiotic ing for such an isotopic "vital effect" cases of photosymbiosis among fossil counterparts, resulting in rapid skel- (Cowen, 1983) and, as mentioned ear- molluscs. etal growth, large size or massive lier, preliminary isotopic investiga- Clinocardium nuttalli is a good ex- structure, high skeleton-to-body ra- tions of modern Tridacna maxima ample of an exception to the conven- tios, or any combination of the three (Jones et al., 1986; Romanek et al., tional environmental criteria. Pho- (Cowen, 1983). This concept, prob- 1987) seemed to support this idea. tosymbiotic C. nuttalli from False ably based on the Tridacna model, Further isotopic investigation of Bay, Washington, occur at a fairly has led to suggestions of photosym- Tridacna maxima (Romanek and high latitude (48-49? N), in a cool, biotic associations for several taxa of Grossman, 1989) cast doubt on the nutrient-rich, tidal flat environment large fossil bivalves, including certain utility of this criterion for identifying consisting of silt and fine sand, with rudists (Kauffman, 1969;Philip, 1972; molluscan photosymbiotic hosts in turbid waters containing consider- Kauffman and Sohl, 1974; Vogel, the fossil record. The isotopic results able organics. Another example from 1975; Skelton, 1979; Kauffman and obtained from Clinocardium nuttalli the same high latitude is Placopecten Johnson, 1988), alatoconchids (Yan- show no evidence of a symbiont-in- magellanicus. Sea scallop popula- cey, 1982), ostreids, and inoceramids duced influence upon either the car- tions with endosymbiotic algae are (Cowen, 1983). However, the case of bon or oxygen isotopic records (Fig. known from deeper, cool waters of Clinocardium nuttalli and several 6). This observation may be ex- coastal Newfoundland in an essen- other modern species suggests that plained, at least partially, by the high tially clastic, nutrient-rich setting the paleontological applicability of ventilation rates for C. nuttalli (Ja- (Naidu, 1971). Other exceptions such the calcification rate criterion is also cobs and Jones, 1989). Meyhofer as Fragum, which burrows in muddy- limited. (1985) reported that C. nuttalli had sand, are discussed in Jones et al. The size-age data presented here the highest pumping rate for its gill (1988). for Clinocardium nuttalli (Fig. 5) size of any of the bivalves tested at clearly indicate that this species does the FHML. This high pumping rate the rapid exchange of 13C- not achieve exceptional size or grow promotes SUMMARY AND CONCLUSION at unusually rapid rates. It does not and 80-depleted respiratory gases, form a massive shell and there is no the incorporation of which is believed Certain populations of Clinocar- evidence to suggest that specimens to cause isotopic disequilibrium (i.e., dium nuttalli from coastal Wash- with symbionts out-calcify their apo- vital effect). The high exchange rates ington and Oregon harbor endosym- symbiotic counterparts. The same could reduce the chances of isotopi- biotic green algae (probably arguments apply to Placopecten ma- cally light carbon being retained in Chlorella) in the siphon, mantle, and gellanicus (Naidu, 1971). In addition the shell carbonate (Jacobs and Jones, occasionally the foot. Infaunal pop- to C. nuttalli, a number of other pho- 1989). Based upon the relatively few ulations consisting of smaller, youn- tosymbiotic bivalves are character- investigations undertaken to date, ger individuals were not infected. The ized by modest size, thin shells, or more isotopic studies on photosym- symbionts were only observed in low to average growth rates, includ- biotic molluscs are clearly warranted. semi-infaunal and epifaunal popu- ing Corculum spp. and Fragum spp. lations which were characterized by These many exceptions do not negate larger and older specimens. These the value of the excessive calcifica- Paleoenvironmental Considerations three mode-of-life groups represent tion criterion in identifying cases of an ontogenetic progression in which photosymbiosis among fossil mol- Conventional wisdom dictates that infaunal clams acquire endosym- luscs. They do, however, indicate that environments with shallow water bionts after the second year of life, paleontologists should not expect a depth and adequate water clarity are whereupon they live semi-infaunally. priori that all cases of molluscan pho- required for symbiont photosynthet- As the intensity of the algal infection tosymbiosis are characterized by en- ic activity. Photosymbiosis is also as- spreads to more tissues, clams be- hanced skeletonization. sociated with tropical, low-latitude, come increasingly epifaunal so that 94 JONES & JACOBS

the largest and oldest individuals are ACKNOWLEDGMENTS Journalof ExperimentalMarine Biology essentially epibenthic. We speculate and Ecology, v. 73, p. 225-242. that this of and We thank Susan Kidwell for sup- JONES, D.S., WILLIAMS,D.F., and ROMANEK, acquisition symbionts C.S., 1986, Life history of symbiont- ontogenetic progression out of the plying the specimens of Protothaca staminea, David Hodell for assis- bearing giant clams from stable isotope substrate by C. nuttalli mimics an profiles:Science, v. 231, p. 46-48. evolutionary trend in which formerly tance with mass spectrometry, Wen- JONES, D.S., WILLIAMS,D.F., and SPERO, infaunal, aposymbiotic bivalves are dy Zomlefer for the illustration in H.J., 1988,More light on photosymbiosis "lured" out of the substrate (evolu- Figure 2, Dave Stetler, Carol Gass- in fossil mollusks: The case of Merce- the benefits of man, S. Herbert, R. Trench, and Lynn naria "tridacnoides": Palaeogeography, tionarily) by photo- Palaeoecology,v. 64, symbiosis (Seilacher, 1984, 1985). It Margulis for help with preparation Palaeoclimatology, and/or interpretation of the TEM p. 141-152. may be that C. nuttalli is in the (evo- KAUFFMAN,E.G., 1969, Form, function and lutionary) process of becoming fully work, and Michael Arthur for his evolution: in MOORE,R.C., ed., Treatise photosymbiotic. comments on an earlier draft of the on Invetebrate Paleontology, Part N, Annual, internal, shell growth manuscript. We are grateful for the Mollusca 6(1), Bivalvia: Geological So- bands a means of determin- thoughtful reviews provided by Dana ciety of Americaand University of Kan- provide sas Press, Lawrence,p. N129-N205. ing age and growth rate. These de- Geary, Ethan Grossman and Chris Romanek. This is KAUFFMAN, E.G., and JOHNSON,C.C., 1988, terminations failed to uncover any paper University The morphologicaland ecological evo- enhancement of the growth rate in of Florida Contribution to Paleo- lution of middle and upper Cretaceous specimens harboring photosymbiotic biology no. 383. reef-buildingrudistids: PALAIOS, v. 3, algae. Stable isotopic analyses of shell p. 194-216. carbonate indicate no distinct isoto- KAUFFMAN,E.G., and SOHL, N.F., 1974, REFERENCES Structure and evolution of Antillean pic differences between mode-of-life Cretaceous rudist frameworks:Natur- groups of C. nuttalli, nor between C. COWEN,R., 1983, Algal symbiosis and its forschendeGesellschaft Basel Verhand- nuttalli and aposymbiotic Prototha- recognitionin the fossil record:in TEV- lungen, v. 84, p. 399-467. ca staminea collected from the same ESZ,M.J.S., and MCCALL,P.L., eds., Bi- KAWAGUTI,S., 1983, The third record of as- locality. Thus, the oxygen and carbon otic Interactions in Recent and Fossil sociation between bivalve mollusks and of C. nuttalli Benthic Communities:Plenum Publish- zooxanthellae:Proceedings of the Japan isotopic compositions ing Co., New York, p. 431-478. Academy, Series B, v. 59, p. 17-20. do not display a symbiont-induced EREZ,J., 1978, Vital effect on stable isotope KRANTZ,D.E., JONES,D.S., and WILLIAMS, fractionation effect. composition in foraminifera and coral D.F., 1984, Growthrates of the sea scal- Several aspects of the molluscan- skeletons: Nature, v. 273, p. 199-202. lop, Placopecten magellanicus, deter- algal association in Clinocardium FRASER,C.M., 1931, Notes on the ecologyof mined from the 180/160 record in shell nuttalli appear to violate conven- the cockle Cardium corbis Martyn: calcite:Biological Bulletin, v. 167,p. 186- tional wisdom such rela- Transactions of the Royal Society of 199. concerning Canada,Section V, p. 59-72. KRANTZ,D.E., WILLIAMS,D.F., and JONES, tionships. These include: lack of a HALLOCK,P., and SCHLAGER,W., 1986, Nu- D.S., 1987, Ecologicaland paleoenviron- specialized morphology to promote trient excess and the demise of coralreefs mental informationusing stable isotope light exposure; absence of a symbi- and carbonate platforms:PALAIOS, v. profiles from living and fossil molluscs: ont-enhanced calcification rate; no 1, p. 389-398. Palaeogeography, Palaeoclimatology, detectable fractionation effect in the HARTMAN, M.C., and PRATT,I., 1976, Infec- Palaeoecology,v. 58, p. 249-266. carbon and oc- tion of the heart cockle, Clinocardium LUTZ,R.A., and RHOADS,D.C., 1980,Growth isotopic composition; nuttalli, fromYaquina Bay, Oregon,with patterns within the molluscan shell: An currence in a high-latitude, cool wa- an endosymbioticalga: Journal of Inver- overview: in RHOADS,D.C., and LUTZ, ter, nutrient-rich, clastic environ- tebrate Pathology, v. 28, p. 291-299. R.A., eds., Skeletal Growth of Aquatic ment. A re-examination of other JACOBS, D.K., and JONES, D.S., 1989, Pho- Organisms:Plenum Publishing Co., New photosymbiont-bearing molluscan tosymbiosis in Clinocardium nuttalli: A York, p. 203-254. taxa reveals that many species, like model for isotopic "vital effects" with MEYHOFER,E., 1985, Comparative pumping C. nuttalli, do not share the char- implicationsfor the fossil recordof pho- rates in suspension-feeding bivalves: tosymbiosis:Geological Society of Amer- Marine Biology, v. 85, p. 137-142. acteristics of conventional wisdom. ica Annual Meeting Abstractswith Pro- MUSCATINE,L., and GREENE,R.W., 1973, While these exceptions do not un- grams, v. 21, p. A77. Chloroplastsand algae as symbionts in dermine the value of the convention- JONES,D.S., 1983, Sclerochronology:Read- molluscs: International Review of Cy- al criteria used by paleontologists to ing the record of the molluscan shell: tology, v. 36, p. 137-169. support hypotheses of photosymbio- AmericanScientist, v. 72, p. 384-391. NAIDU, K.S., 1971, Infection of the giant that cases of JONES, D.S., ARTHUR,M.A., and ALLARD, scallop Placopecten magellanicus from sis, they suggest many D.J., 1989, Sclerochronologicalrecords Newfoundland with an endozoic alga: molluscan photosymbiosis in the fos- of temperature and growth from shells Journalof InvertebratePathology, v. 17, sil record presently go unrecognized. of Mercenaria mercenaria from Narra- p. 145-157. Paleontologists are urged to develop gansett Bay, Rhode Island: Marine Bi- NAIDU, K.S., and SOUTH, G.R., 1970, Oc- better criteria for the recognition of ology, v. 102, p. 225-234. currence of an endozoic alga in the giant molluscan paleophotosymbiosis and JONES, D.S., WILLIAMS,D.F., and ARTHUR, scallop Placopecten magellanicus Growth and Canadian Journal of are cautioned against dismissing pos- M.A., 1983, history ecology (Gmelin): Zoology, of the Atlantic surf clam, Spisula soli- v. 48, p. 183-185. sible cases because they do not con- dissima (Dillwyn), as revealed by stable PHILIP, J., 1972, Paleoecologie des forma- form to conventional criteria. isotopes and annual shell increments: tions a rudistes du Cretace superier- PHOTOSYMBIOSISIN FOSSIL MOLLUSCS 95

l'example du sud-est de la France: Pa- a radiolitid rudistid bivalve and its im- Further implications of a model for sta- laeogeography, Palaeoclimatology, plication of marginal mantle feeding in ble isotopic fractionationby scleractini- Palaeoecology, v. 12, p. 205-222. the group:Paleobiology, v. 5, p. 90-106. an corals: Geochimica et Cosmochimica ROMANEK,C.S., and GROSSMAN,E.L., 1989, SWART,P.K., 1983, Carbon and oxygen iso- Acta, v. 40, p. 34-39. Stable isotope profiles of Tridacna max- tope fractionation in scleractinian corals: WEBER,J.N., and WOODHEAD,M.J., 1970, ima as environmental indicators: PA- A review: Earth Science Reviews, v. 19, Carbon and oxygen isotope fractionation LAIOS,v. 4, p. 402-413. p. 51-80. in the skeletal carbonate of reef-building ROMANEK,C.S., JONES,D.S., WILLIAMS,D.F., TRENCH, R.K., 1979, The cell biology of corals: Chemical Geology, v. 6, p. 93-117. KRANTZ,D.E., and RADTKE,R.L., 1987, plant-animal symbiosis:Annual Review WILLIAMS,D.F., SOMMER,M.A., and BENDER, Stable isotopic investigation of physio- of Plant Physiology, v. 30, p. 485-531. M.L., 1977, Carbon isotopic composi- logical and environmental changes re- TRENCH,R.K., WETHEY,D.S., and PORTER, tions of recent planktonic foraminifera corded in shell carbonatefrom the giant J.W., 1981, Observationson the symbi- of the Indian Ocean: Earth and Plane- clam, Tridacna maxima: Marine Biolo- osis with zooxanthellae among the Tri- tary Science Letters, v. 36, p. 391-403. gy, v. 94, p. 385-393. dacnidae (Mollusca, Bivalvia): Biologi- WILLIAMS,D.F., ROTGER, R., SCHMALJOHANN, SEILACHER,A., 1984, Constructional mor- cal Bulletin, v. 161, p. 180-198. R., and KEIGWIN,L., 1981, Oxygen and phology of bivalves: Evolutionary path- VOGEL, K., 1975, Endosymbiotic algae in carbon isotopic fractionation and algal ways in primary versus secondary soft- rudists?: Palaeogeography, Palaeocli- symbiosis in the benthic foraminiferan bottom dwellers:Palaeontology, v. 27, p. matology, Palaeoecology, v. 17, p. 327- Heterostegina depressa: Palaeogeogra- 207-237. 332. phy, Palaeoclimatology, Palaeoecology, SEILACHER,A., 1985, Bivalve morphology and WATSON, M.E., and SIGNOR,P.W., 1986, How v. 33, p. 231-251. function:in BROADHEAD,T.W., ed., Mol- a clam builds windows: Shell microstruc- YANCEY,T.E., 1982, The alatoconchid bi- lusks, Notes for a Short Course, orga- ture in Corculum (Bivalvia: Cardiidae): valves: Permian analogs of modern tri- nized by D.J. BOTTJER,C.S. HICKMAN, Veliger, v. 28, p. 348-355. dacnid clams: Proceedings of the Third and P.D. WARD, University of Tennes- WEBER, J.N., DEINES,P., WEBER,P.H., and North American Paleontological Con- see, Knoxville, Department of Geological BAKER,P.A., 1976, Depth related changes vention, v. 2, p. 589-592. Sciences: Studies in Geology, v. 13, p. 88- in the '3C/'2Cratio of skeletal carbonate 101. deposited by the Caribbean reef-frame ACCEPTED AUGUST 1, 1991 SKELTON,P.W., 1979, Preserved ligament in building coral Montastrea annularis: