J. Adv. Mar. Sci. Tech. Soci. Vol.1. No.l. 1995. PP. 11-21

Growth variation of elemental concentrations in the calcitic bivalve shell of

Mizuhopecten yessoensis (Jay) from Mutsu Bay, Northern Japan

mm m±mFujio Masuda *

Abstract

The variation of concentrations of elements with growth in the calcitic outer shell layer of Mizuhopectenyessoensis (Jay), living in Mutsu Bay, northernJapan was for the first time presentedin detail.The variationwas clarifiedfor the contentsof elevenelements: Li, Na, K, Mg, Sr, Ba, Fe, Mn, Al, B and P.The variation patterns make to divide the elements into the following four groups: (1) magnesium,

strontium, barium, iron and manganese; showing the seasonal variation pattern. These are occupied in thelattice sites in thecalcium carbonte's crystal. (2) lithium and sodium; showing the seasonal variation, patternwith reverse pattern for the former group. These are associated with inclusions and/or organic matrices. (3)aluminium, pottasium and phosphorous; showing a random variation pattern. These are originated fromforeign detrital grains. (4) boron; showing no seasonal variation pattern and partially high content Boronmay be exsited as an other phase with small garains.The difference among these variation patterns originates in the difference of phase within shell. The

variations of thefirstand secondgroupscan beexplained by "optimum condition model" proposed herein.

1. Introduction

Calciumcarbonate is the mostcommonly occurring skeleton of marineorganisms, such as mollusca, coraland foraminifera. Chemical compositions in modem and fossil skeleton of calcium carbonate have beenstudiedas a tool of paleoenvironmental reconstruction. From these results many empirical relationsbetweenchemicalcomposition of modem skeletonsand environmental conditions (mostly temperatureand salinity)have been propose (Chave, 1954; Pilkey and Hower, 1960; Dodd,1965; Hallam and Price, 1968; Masuda,1976; Kolesa, 1978). These empirical relations were reviewed by Milliman (1974) and Masuda(198la, 1984). Consequently, both positive and negative correlations are present among these empiricalrelations, whilein somecases no correlation wasfound. Masuda (1984) proposed a model for the purposeofgiving an explanation to the complex phenomena concerning these empirical relations. The mechanism ofthis phenomena, however, is not' well understood. To understand the mechanism for these complicatedphenomena, we must get the detailed elemental compositions, distribution and variations within the variousskeletons of calcium carbonate.

In this paper, the variation of elemental concentrations with growth in the calcitic shell of the bivalve,Mizuhopecten yessoensis with definite career is for the first time presented in detail and is discussedconceming shell growth and environmental variation.

•DepartmentofEarth and SpaceScience,FacultyofScience, OsakaUniversity, ToyonakaOsaka560,Japan. 3MJz¥M¥t$

2. Material and method

Mizuhopecten yessoensisTwo shells of Mizuhopecten yessoensis (Jay), collected from Moura, Hiranai-machi in Mutsu Bay, Aomori

Prefecture in 1973, areselected for thepresent study. Theshell (A) is9.77centimeter in length, 9.35centimeterin height, and 2.30 centimeter in thickness. The shell (B) is 13.12 centimeter in length, 12.27 centimeter inheight, and 2.25 centimeterin thickness.The shell of Mizuhopecten yessoensis is divided into the two layers, outer and inner shell layers. The

outer shell layer of right valve is used for the analysis of seasonal variation. It is composed of the calcite,Cluster IV divided by Masuda (1987), and has the foliated shell structure (Kobayashi, 1971).Several distinct rings, which indicated the stages of very slow growth or pause of growth, appear on theshellsurface (Fig. 1). The periods of forming of therings andannual variation of the shell growth are ableto be decided, as follows, by aquaculture data (Kanno et al., 1974; Yamamoto, 1976; Ito et al., 1976).The youngest ring is a disturbance ring formed by liberationat late March to early April.The other prominent rings in older stage are so-called "annual rings". The annual ring of Mizuhopecten

yessoensis in the Mutsu Bay is formed during late summer to autumn, from the end of August to the endof September (or the beginning of October) which is the warm period of bottom-seawater (higher than20°C). Accordingly, thedistinct annual rings of theshell are"summer rings".There is a weak ring between two summer rings, as observed on the shell surface of the second and third

years (Fig. 1). This is the "winterring" which is formed during late winter to early spring (February toMarch), the cool periodof bottom-seawater (lowerthan 5 °C).During the first summer to autumn, the shell growth is slow but continuous and the annual ring is not

formedor weak. In the older age, the shell growsonly at the optimumconditionand the prominentannualring is consequently formed. Fast-growing periods, corresponding to the optimum condition, are fromautumn to early winter and from late spring to summer.Figure 1 also presents the life calendar of Mizuhopecten yessoensis with the annual changes of

temperature and salinity (average of quarter a mouth from 1965 to 1977) recorded by Aquaculture Center,Aomori Prefecture at Moura in Mutsu Bay.

AnalyticalprocedureA series of samples, each weighing several milligrams, is carefully collected from the cleaned outer layer,

from ventral margin to umbo, using a dental drill under the microscope. Fifty-one samples from the shell (A)and one hundred and one samplesfrom shell (B) could be separatedas shown in Fig. 1.The samples are analyzed by an inductively coupled plasma-opticalemission spectrometer. The method has

been described elsewhere (Hirano et al., 1979). The method gives the simultaneous determination of elevenelements, such as Ii, Na, K, Mg, Sr, Ba, Fe, Mn, Al, B and P.

3. Results

The variations of contentsof the elevenelementsare illustrated in Fig. 1. Both shells (A) and (B) show thesimilar pattern of variation.The mean value and range betwen minimum and maximum values (in ppm) are as follows: Shell (A):

Li=3.2, 0.68-5.0; Na=4400, 3250-5750; K=240, 145-590, Mg=840, 545-1900; Sr=1050, 990-1150; Ba=3.3,1.0-10.0; Fe=78,13-500; Mn=12, 6.0-57; Al=33,2.4-93; B=46,2.2-250; P=370,260-430. Shell (B): Li=6.5,1.7-58; Na=4450, 2800-5700; K=280, 160-540; Mg=760, 450-2650; Sr=1060, 990-1150; Ba=2.1, 0.7-20;Fe=54,11-540; Mn=8.7,5.4-31; Al=44,15-220; B=30,3.3-340; P=400,320450.

12

320-200i

BOTTOM

TEMPERATURE10

5cmi I l i I

'•J '-./ salinity V 33.4

'a's'o"Vd;j'f'm'am'j j a's'o'n'dIj'f'm'a'm'j j asondIjf'

Fig. I Growth variation of elemental composition withdistance fromthe umbo in the outer calciticshell layer of Mizuhopecten yessoensis from Mutsu Bay. The lifccalendar of the shell andannualtemperature and salinityof bottom-seawater (ten-year average)are also presented.

13

4. Discussion

Division of the variation patternsThe variation patterns of theelements shownin Fig. 1 canbe divided intofourgroups.(1) magnesium, manganese, iron, strontium and barium; these show the "seasonal" variation pattern with

higher concentrations near the annual (summer) ring. In the older shell these elements roughly show theincreasing concentrations. The phenomenon can be called "age effect". The increasing concentrations,possibly an age effect, of bariumand strontium werepreviously found in the older shell of aragonitic freshwater bivalve (Carell et al., 1987).These can be subdivided into three as follows: (1-a) magnesium contents show the typical "seasonal"

variation pattern; (1-b)manganese, iron and barium contents show the "semi-annual" variation patternwiththe peak of higher concentration at the both of "annual"(summer) and winter rings; and (1-c) strontiumcontents show the "annual" variation pattern with noised peaks. The variation pattern of strontium at theyounger age (firstandearlysecondyear)is similar to thatof thesodium and lithium, mentioned next

(2) lithium and sodium: Their concentrations also show the "seasonal" variation pattern with the lowercontent near the "annualring". A negativerelationexistsbetween the contents of first group (e.g., Mg. Mn)and second group (e.g., Li, Na).

(3) aluminium, pottasium and phosphorous: Their concentrations show the "random" variation patternwithout any seasonal effect. The variation pattern of pottasium within the shell (A) is similarto that of thefirstgroup(e.g., Mg). However, the pattern within the shell (B) doesnot showthatof the firstgroup. Thevariation pattern of phosphorous alsoslightly resembles to thatof strontium.

(4) boron: The concentration of boron shows no "seasonal" variation pattern with partial high peaks. Mostparts withinshell show very low concentration.

PC-IR diagramIt is necessary to introduce the crystalstructure control on traceelementpartition in shellformation basedonthe partition coefficient-ionic radius (PC-IR) diagrams presented by Onuma et al. (1979) and Masuda andHirano (1980). The reason is that the PC-IR diagrams are already distinguishable into three groups ofelements, such as, divalent alkaline earth elements, transition elements and monovalent alkali metal elements.Figure 2(A) is a PC-IR diagram for the shellof Mizuhopecten yessoensis - extrapallial fluid system. The

extrapallial fluid is the mineralizing solution of shell formation, and secreted from the outer mantleepithelium. It is clearly isolated by thecellmembrances from theenvironmental waterin whichmolluscs areliving. The diagram is drawn based on the data of mean values of content of the outer shell layer ofMizuhopecten yessoensis from Mutsu Bay (Masuda and Hirano, 1980) , extrapallial fluid (Wada andFujinuki, 1976), and ionic radii (Shannon, 1976). Unfortunately, as the chemical composition of theextrapallial fluid of Mizuhopecten yessoensis is unknown, that of Chlamys nobilis, whichhas a same shellstructure (foliated structure) as Mizuhopecten yessoensis,is used. These data are given in Table 1.

Table.1Elementalabundances and partition coeffients

Elements y Na K Ni Mg Cu Zn Fe Mn Ca Sr Ba Al Cr V Sc

Contents(ppm)

(DShell 4.1 3760 170 0.2 690 0.45 16 41 12 3810000 1060 4.7 24 1.4 0.06 0.006

(2)Extrapallial fluid 0.19 9780 428 1183 0.09 1.05 0.75 0.08 398 8.1

(3)Seawater 0.17 10500 380 0.002 1350 0.003 0.01 0.01 0.002 400 8 0.03 0.01 0.00005 0.002 0.000004

Electriccharge 1+ 1+ 1+ 2+ 2+ 2+ 2+ 2+ 2+ 2+ 2+ 2+ 3+ 3+ 3+ 3+

Ionic radius(A)* 0.76 1.02 1.38 0.69 0.72 0.73 0.74 0.78 0.83 1.00 1.18 1.36 0.53 0.615 0.64 0.745

Partition coefficients(Log.)

(iy(2) 1.33 -0.42 -0.40 -0.23 0.70 1.18 1.74 2.17 2.98 2.12

(1V(3) 1.38 -0.45 -0.35 2.00 -0.29 2.17 3.20 3.61 3.78 2.98 2.12 2.19 3.38 4.45 1.48 2.18

(1) Mean values of shell outer shell layer ofMizuhopectenyessoensis from Mutsu Bay (Masauda and Hirano, 1980)(2) Extrapallial fluid of Chlamysnobilis (Wada and Fujinuki, 1976)(3) Mean values of seawater (Hill, 1963) *Ionic radius (Shannon, 1976)

14

Fig. 2

0.8 1.0 1.2

IONIC RADIUS (A)

10

- 10 -

(B)"*Cr '1 T —I r— i -i r

" Mn •

Al 9Te

-

yzn Ca-

Nij/Cu /'Oft /

Sc/\sr Ba

o

Va 'PLi

"

-

Mg

i i > •

aNa

• < •

Ka

•

io2-

- 10-

- 10°-

1410

0.6 0.8 1.0 1.2

IONIC RADIUS (A)(A) Partition coefficient-ionic radius diagram for the outer shell layer ofMizuhopecten yessoensis and the extrapallialfluid of Chlamys nobilis, (B) Partition coefficient-ionic radius diagram for the outer shell layer of Mizuhopectenyessoensis and seawater. Data are shown in Table 1.

1.4

The partition coefficient is defined as the ratio of concentrationof an element in shell for fluid or water.Divalent ions, such as alkaline earths and transition elements (Mg, Ca, Sr, Cu, Zn, Fe and Mn), make a

parabolic curve with a peak at the point for calcium in Fig. 2(A). The parabolic curve means that theelemental partition of the divalent ions inthe shell - extrapallial fluid system isprimarily determined by thecrystal structure ofcalcium carbonate, which offers a place accommodating a cation, such as calcium. Thatis, thecrystal structure control forunfitted elements is presented as parabolic curve and these divalent ionsoccupyregularly the latticesitesin crystal structure of calcium carbonate.

Monovalent ions, such asalkali metal elements (Na, Li and K), don't lie onthe parabolic curve suggestingthe absence oftendency for the alkali metal elements tooccupy the lattice site ofthe crystal structure.Fig. 2(B) isa PC-IR diagram for the shell ofMizuhopecten yessoensis - seawater system. The mean values

of theshell(Wadaand Fujinuki, 1976) andseawater (Hill, 1963) areshownin Table 1.The contents ofmagnesium, calcium and strontium ofextrapallial fluid are nearly equal tothose ofseawater(Table 1). Accordingly, on the PC-IR diagram ofFig. 2(B) the same shaped parabolic curve shown in Fig.2(A) can be drawn from the values ofalkali earth elements, magnesium to strontium, with a peak located atthe point of calcium. This means that for the concentration of alkali earth elements the crystal structurecontrol is an important factorthroughout shellformation.Theothershifted curve is abletodraw forthedivalent transition elements, Mn,Fe,Zn,CuandNi,which areplotted above the left side of the parabolic curve ofalkaline earth elements onthe diagram (Fig. 2(B)). Theselarge values of the partition coefficients of transition elements are derived from the concentration effectduring the interchange from seawater to extrapallial fluid. Namely the transition elements are mainlyconcentrated due to thephysiological process during the inter change from seawater toextrapallial fluid, andare partitioned regularly in crystal structure by the physiochemical process during the interchange fromextrapallial fluid to shell.

Monovalent alkali metal elements (Li, Na and K) and trivalent elements (Al, Cr, Cand Sc) inFig. 2(B)showno parabolic curve,notcontrolled by crystal stmcture.

15

Four groups of elements and Optimum condition modelThe variation patterns presented in thispaperare discussed for each groupof elements basedon the above

mentioned results concluded from the PC-IR diagrams.(1) Magnesium, manganese, iron, strontium andbarium(1-a) Magnesium : The typical "seasonal" variation of magnesium content was explained by the crystal

growth kinetics combined with metabolic processes (Masuda, 1985). The reason why it shows the typicalcyclic pattern is that the calcitic shell of Mizuhopecten yessoensis has a higher magnesium content thanaragonitic shell, andthatmagnesium ions regularly occupy thelattice sites in thecrystal structure of calciumcarbonate of the shell, as mentioned before. It is considered that the crystal growth kinetics, combined withmetabolic processes, directly give an influence formagnesium content within shells. Therefore, theintensityof calciumcarbonatemetabolism is mainlycontrolled by environmental conditions, such as temperature andsalinity.In thiscase,the variation of magnesium content couldbe mainly explained by theoptimum condition model(Masuda, 1981a, b,1984) basedon the intensity variation of calcium carbonate metabolism controlled by theseasonal change of seawater temperature and"age effect". Seawater temperature isoneof themost importantenvironmental factors known to influence to mollusca. Then, this model is constructed under the hypothesisthatMizuhopecten yessoensis has the optimum temperature for calcium carbonate metabolism, and that theintensity of themetabolism decreases toward lower or higher values of temperature from theoptimum one.Thechange of theintensity shows the "normal distribution-like curve", given in Fig. 3(C). Moreover, themetabolism has an "ageeffect", higher metabolism with strong intensity and wide rangeof metabolismactivity inyounger age, and lower metabolism inolder age (shown in Fig. 3(C)). Intensity ofthe metabolismdirectly correlates withtheelemental concentration, asdiscussed later.

In Fig. 3(A), the variation of magnesium content (note the higher content plot on the lowerpart) isrearranged against annual time estimated by the growthrings and the lowest content in a cycle. Thechange of annual temperature of bottom seawater is also presented in Fig. 3(D). In this model theactivity range of metabolism for temperature changes from wide to narrow with ageas shown in Fig.3(D). Therefore, the period of pause ("unsuitable" condition) for metabolism represented by "annualring" increases year by year with age as given as shadow bands in Fig. 3(D). Based on theoptimumcondition model shown in Fig. 3(C) and (D), the variation of the intensity of calcium carbonatemetabolism can be calculated from setting up the followingcondition; the optimum temperature is 12-13°C and the range of active metabolism in from 5 to 20°C. Namely the intensities of calciumcarbonate metabolism can be plotted for the variation of temperature of Fig. 3(D) using threeconditions of calcium carbonate metabolism as shown in Fig. 3(C). The obtained variation of thisintensity in Fig. 3(B) is similar to the variation of magnesium contents (Fig 3(a)). This relationbetween the intensity of calcium carbonate metabolism and magnesium contents can be explained asfollows.

The shell precipitates "pure" crystals contained lower magnesium contents at the optimum condition ofcalciumcarbonate metabolism. The "pure" crystals meansa narrow parabolaon the PC-IR diagramforthe shell andextrapallial fluid system (Fig. 2(A)). Thenarrow parabola indicates thecondition in nearlyequilibrium. Under this condition, the strong selection of unfitted elements in the crystal structure ofcalcium carbonate, makes- "pure" crystals with lower magnesium content On the contrary, the lowerintensity of calcium carbonate metabolism, giving a broad parabola, leadsthe highermagnesium contentin "impure" crystals. Actually, parabolic curvefor a partof the lowermagnesium content(highCaC03metabolism) arenarrower thanthatat thehigher content (low CaC03 metabolism) on thePC-IR diagramof the shells (A) and (B) as shown in Fig. 4. (1-b) Manganese, iron and barium : The optimumtemperature model proposed above canalsoexplain roughly thevariation of manganese, ironandbariumcontents shown in Fig. 1. As shown in Fig. 4, the manganese and iron have the same behavior asmagnesium.

16

£,o3LBNX

u

oo

o10

<10°-

100.6

(A)500

Growth >MAM J J ASONDjJ FMAM J J ASONDiJ FMAM J J ASONDjJFMA

1AMJ J ASONDiJ FMAMJ J A SONDiJ FMAMJ J ASONDiJ FMA

Fig.3 Variation pattern of magnesium contents (A) can be simulatedby the variation of intensity of calciumcarbonatemetabolism(B). The variation of intensity is calculated for the variation ofannual temperature (D) using by the three age effectedconditions of calcium carbonate metabolism (C). This can becalledas optimumtemperature model.

—i 1 1 1 1 1 1 r

Shell (A) o •Shell (B) a m

Mn

4/8 'sII

II

u

Mgf

8I

Na

f

B

0.8 1.0 1.2 14

IONIC RADIUS (A)

Fig.4 Partition coefficient -ionic radius daiagram for the shellof "optimum" condition (low magnesium content, opensymbols) and the shellof "unsuitable" condition (highmagnesium content, solid symbols) - the extrapallialfluidof Chlamys nobilissystem.

Transition elements of manganese andiron aremainly controlled bycrystal stracture during thelaststep of shellfoiming system, asmentioned before. The last step means the elemental partition between shell and extrapallialfluid (Fig. 2(A)). However, most of concentration of divalent transition elements arecarried by thecomplexphysiological activities from seawater toextrapallial fluid (Rg.2(B)). Moreover, some ofthe transition elementsmay interact with physiological control as "bio-elements", because the calcium carbonate crystals of shell areenclosed andinterbedded by thethin (several fi to several tens fx)organic sheets of proteinaceous compoundsmatrix, and the transition elements may be concentrated or deconcentrated at the organic matrix. These arereflected bythenoise inthecyclic pattern of manganese andiron.The variation ofmanganese content shows more cyclic pattern than that of iron content Namely, the pattern ofiron has some sharp peaks asnoise. This can beunderstood bythe following fact that the manganese which hasthesimilar ionic radius tocalcium, a major element inshell, iscontrolled toa stronger degree, thanironwhich hassmaller ionic radius thanmanganese bythecrystal structure ofcalcium carbonate.The variation pattern of barium, which is similar to manganese and iron, means that barium has the same

behavior as thedivalent transition elements (e.g., Mg. Fe). This maybe supported by thefollowing work. OnthePC-IR diagram for shell - seawater system(Fig. 2(B)), a large parabolic curve can bedrawn from the points ofdivalent transition elements, such asnickel, copper, zinc, iron andmanganese, tothepoint of this barium above aparabolic curve through the points ofmagnesium, calcium and strontium. (1-c) Strontium : The variation ofstrontium concentration follows a similar pattern to thatof magnesium. This is because both strontium andmagnesium elements occupy the lattice sites in the crystal structure ofcalcium carbonate based onthe PC-IRdiagrams ofFig. 2,as mentioned before. Infact, as shown inFig. 4,the part oflower (higher) magnesium contentprecipitated atthe "optimum" ("unsuitable") tenirjerature shows the lower (higher) strontium contentHowever, the pattern of strontium at the first to second year is different from that of magnesium. The

phenomenon cannot beexplained here. An inverse relation isfound weakly between the pattern ofstrontium andmagnesium. Moreover, there are sharp peaks between "annual" (summer) and winter rings which are notrecognized inthe pattern ofmagnesium. The strontium content inshell may beaffected not only bytemperaturebut salinity whose annual variations are shown in Rg. 1. It may bedifficult to understand the environmentaleffectof strontium contentin the calcitic shellas Masudaand Hirano(1980) showedthat the calcitic shell had alower strontium content (mean, 890 ppm) than aragonic shell (average, 1950 ppm), and the variation range ofthevalues is small (160ppm). This isconsistent with the fact that the seasonal variation pattern ofstrontium inthearagonitic shell ofMeretrix lusoria ismore pronounced than that for magnesium (Masuda, 1985).(2) Lithium andsodiumThe variations of lithium and sodium contents which show the typical cyclic pattern are controlled by the

absolute amounts of inclusions and/or organic matrices within shell. This canbe deduced from thefollowingresults.

Namely, monovalent alkali metal elements, such aslithium and sodium, have notendency tooccupy the laticesite ofcrystal structure ofcalcium carbonates, because the elements do not lie onthe parabolic curve inthe PC-IRdiagram (Rg. 2(A)). Isomorphic substitution ofmonovalent elements incalcium carbonate ishardly expected,because such substitution requires coupled substitution which involves the water-soluble components). Thedistribution of the elements should be treated separately as inclusions suchas carbonates and/or amorphouscompounds. Actually, much trapped water and many organic material inclusions were reported within the shellofMytilus califomianus (Towe, 1972). Moreover, the sodium/calcium variations and their absolute amounts incalcitic shell ofMytilus edulis were best explained byassuming that a substantial amount ofsodium was adsorbedon thecalcium carbonate crystal surface (Lorens andBender, 1980).

The following relation can beobtained among the amount of organic matrix and inclusion, crystal size andgrowth rate (Wada, 1972; Iwasaki, 1977); (a) increase ofgrowth rate = small crystal = rich organic matricesand inclusions; (b) decrease of growth rate = large crystals = poor organic matrices and inclusions. Theobtained result thevariation patterns ofmagnesium andsodium or lithium show a negative relation, concludedthattherelation (a)means higher condition ofcalcium carbonate metabolism, ontheother hand, therelation (b)means lower conditionof calcium carbonatemetabolism. Consequently, the relations can be rewritten as

18

follows: (a) high calcium carbonate metabolism = "optimum" condition = fast growth rate = small and "pure"crystals = richorganicmatricies and inclusions, (b) lowcalciumcarbonate metabolism = "unsuitable" condition= low growth rate = largeand "impure" crystals = poororganic matrices and inclusions. The condition ofcalcium carbonate metabolism is controlled by the environmental condition, such as temperature and salinity.According totheoptimum condition model proposed before, theoptimum value isnotnecessary toagree to thehighest (orlowest) values of temperature or salinity. Moreover, theoptimum value is thecharacteristic of eachspecies. Therefore, this supports the view that there is nolinear correlation or there is both of the positive andnegative correlations between concentration of trace element of marine carbonate and water temperature orsalinity forsomeempirical relations aspreviously reported by Masuda (1984).(3) Aluminium, pottasium andphosphorousThe variation patterns ofaluminium, pottasium and phosphorous can make nosystematic interpretation. They

show partially seasonal variation pattern. The presence ofthese elements are not controlled bycrystal structureof calcium carbonates. The presence of aluminium has been considered theprovenance from the foreigndetrital grains (Millirnan, 1974; Masuda and Hirano, 1980). The variation pattern of pottasium is partiallysimilar tothe divalent transition elements. Pottasium issame group with sodium onPC-IR diagram. Wecan notexplain the presence ofpottasium, but itmay bewith organic matters and detrital grains. Phosphorous ishighlyconcentrated from seawater to extrapallial fluid (Wada and Fujinuki, 1976). Phosphorous may also be withinclusions, organic matrices anddetrital garains.(4) Boron

The variation pattern ofboron shows a flat ofvery low concentration with partial sharp ofhigh concentration.The high concentration ofboron correspond to that ofaluminium (Rg. 1). The pattern strongly indicates thattheelement maybeexisted assmall grains ofotherphase with terrigenous contarnination.

5. Conclusions

Growth variation of elemental composition in a calcitic bivalve shell was for the first time presented indetail. Twoseries of samples, onehundred andoneand fifty-one, were collected from theouter shell-layerof Mizuhopecten yessoensis (Jay), taken from Mutsu Bay, northern Japan. The variation of the two serieswereclarified for thecontents of eleven elements, Li, Na,K, Mg,Sr,Ba, Fe, Mn,Al, B and P. The variationpatterns maketo dividetheelements intothefollowing fourgroups.(1) Magnesium, strontium, barium, iron and manganese: Their contents show the seasonal variation patternwith higher content near the annual ring. Particularly, the magnesium content shows the typical cyclicvariation. These variation patterns can beexplained by the optimum temperature model onthe crystal growthkinetics, combined with calcium carbonate metabolism. These elements are controlled by the crystalstructure of calcium carbonate.

(2) Lithium and sodium: Their contents show the seasonal variations are controlled by amounts ofinclusions and/or organic matrices, andnotcontrolled bythecrystal structures.(3) Aluminium, pottasium and phosphorus: Their contents show the random variation pattern and partiallyweak seasonal variation. These elements may beoriginated from foreign detrital grains and organic matrices.(4) Boron: It shows noseasonal variation ofcontents and has low content butpartially high. Boron may beexistedas otherphasewithsmallgrains fromterrigeneous contamination.The following relations canbeobtained: (a) high calcium carbonate metabolism = "optimum" condition =

increase of growth rate =small and "pure" crystals = rich organic matrices and inclusions, (b) low calciumcarbonate metabolism = "unstable" condition = decrease ofgrowth rate = large and "impure" crystals =poororganic matrices and inclusions.

19

References

Carell, B., Forberg, S., Grundelius, E., Henrikson,L., Johnels, A., Lindh, U., Mutvei, H., Olsson, M., Svaristrom, K. and

Westermark, T., 1987. Can mussel shell reveal environmental history? Jour. HumanEnviron., Royal SwedishAcad.ScL(AMBIO), 16,2-10.

Chave, K.E., 1954. Aspect of the biogeochemistry of magnesium. Jour. Geol., 62, 266-283,587-599.

Dodd, J.R., 1965.Environmentalcontrol of strontiumand magnesiuminMytilus. Geochim. Cosmochim. Acta, 29,383-398.

Hallam, A. and Price, N.B., 1968. Environmental and biochemical control of strontium in shells of Cardium edule.

Geochim, Cosmochim. Acta, 32,319-328.

Hill, M.N., 1963.Thesea. IntersciencePublishers,2,554p. John Wiley and Sons, New York.

Hirano, M., Onuma, N. and Masuda, F.,1979. Simultaneous determination of trace elements in shells by inductively

coupled plasma-optical emission spectrometry. Bunseki-Kagaku, 28,313-318.

Ito, S., Nishikawa, N„ Fuji, A., Yamamoto, G. and Mori, K., 1976.The evolution of aquaculture of Pecten yessoensisJay-Technique -. InImai,T. ed, Shallow marine through culture. 211-276, Koseisha-Koseikaku, Tokyo.

Iwasaki,Y., 1977.Shell growth and growth-lines.Kumamoto Jour., ScL (Geol), 10,41-54.

Kanno, H. and Ito, S. et al., 1974.Survey of the resourcesof Pectenyessoensis.Ann.Rep., Aquaculture Center, AomoriPrefecture, 3,68-76.

Kobayashi, I., 1971.Internal shell micro-structure of Recentbivalvian molluscs. ScL Re., NiigataUniv., (E),2,27-50.

Kolesa,P.T., 1978.Magnesium in calcite from a corallinealgae.Jour. Sediment. Petrol., 48,815-820.

Lorens, R.B. and Bender, M.L., 1980. The impact of solution chemistry on Mytilus edulis calcite and aragonite.Geochim. Cosmochim. Acta, 44,1265-1278.

Masuda, F., 1976. Strontium contents in shell of Glycymeris. Geol Soc. Japan,82,565-572.

Masuda, F.,198la. Can the elemental contents in skeletal carbonates be used as Paleo-thermometer?Jour. Geography,

95,1-26.

Masuda, F., 1981b.Variationof Mg and Na contentswith growth in the outer shell layer of Mizuhopecten yessoensisfrom Mutsu Bay, northernJapan.Ann. Rep.Jnst. GeoscL, Univ. Tsukuba, 7,49-51.

Masuda, P., 1984. How to understand the empirical relations between elemental content of skeletal marine carbonatesand watertemperature or salinity. Ann. Rep., Inst. GeoscL, Univ. Tsukuba, 10,92-95.

Masuda,F., 1985. Seasonal growthvariation of elemental contents in the bivalveMeretrix lusoria fromthe AriakeSea,Kyushu, westernJapan.Ann. Rep.Jnst GeoscL, Univ. Tsukuba, 11,35-40.

Masuda, F., 1987.Four groups of skeletalcalcite and aragoniteby Sr-Mg systematics.Awi.Rep., Inst. GeoscL, Univ.Tsukuba, 13,96-99.

Masuda, F. and Hirano, M., 1980. Chemical composition of some modern marine pelecypod shells. ScL Rep.Jnst.GeoscL, Univ. Tsukuba,Sec. B, 1,163-177.

Milliman, J.D., 1974. Marinecarbonates.375p. Springer-Verlag, Berlin.

Onuma, N., Masuda, F., Hirano, M. and Wada, K., 1979.Crystal structurecontrol on trace element partition in molluscan

shell formation. Geochim. Jour., 13,187-189.

Pilkey, O.H.andHower,J., 1960. The effectof environment on theconcentration of skeletal magnesium and strontiumin Dendraster. Jour. GeoL, 68., 203-216.

Shannon, R.D., 1976. Revised effective ionic radii and systematics studies of interatomic distances in halides and

chalcogenides.Acta Cryst.,A, 32, 751-767.

Towe, K.M., 1972. The structure of some bivalve shell carbonates prepared by Ion-beam thinning. Calc. Tiss.Res., 10,

38-48.

20

Wada, K., 1972.Nucleadon and growth of aragonitecrystals in the nacre of some bivalve mollsucs.Biomineralizjation, 6,141-159.

Wada, K. and Fujinuki, T., 1976. Biomineralization in bivalve molluscs with emphasis on the chemical composition ofthe extrapallial fluid. In Watabe, N. and Wilbur, K.M. eds., Themechanism ofmineralization in the invertebrates andplants. 175-190, Univ. South Carolina Press.

Yamamoto,G., 1976.The evolution of aquacultureof PectenyessoensisJay- Biological study-. In Imai, T. ed., Shallowmarine through culture. 191-210, Koseisha-Koseikaku, Tokyo.

9 t

tz. &&'**->frZ4#'Ji'-'7lZ%tte>tiZ. (l)Mg. Sr, Ba, Fe, Mn : ^IM&«hW»£

&+. mmgtmm%mffl%W)*7Ft. z.tLt>\ffi8Mmm~, Ba, Fe, Mni&m&Ri-t&tE•*-*. (2)u, Na: wm&^ffimmzTfii-. %*i\z(\)7)i—'7<o$tmb\m^9-yttzz. ^nmm^'mmKWo. (soai, k, p:f»«t*^s4o. M^Kimo&tiitzftxn?, payyWtf&bLXfttetz. (4)B:$OTa»tt3sS&i\ IWR^fcL-r-apirMfti-ft. M<rW&M.

tit* rv^v^ (igo) ^n*^s/«>Af^w=av^ »>) rt«=/hs& (*£&) ^0hb=^m (^a)

Recieved 14July 1994Reviced 20 December 1994

Accepted 7 January 1995