Biomineralization features in scleractinian coral skeletons: source of new taxonomic criteria Caracteristicas de la biomineralizacion en los esqueletos de corales escleractinios: fuente de nuevos criterios taxonomicos Jean-Pierre Cuif, Yannicke Dauphin and Pascale Gautret Facalte des Sciences, France - Bat 504 Geologie U.R.A. C.N.R.S. 723 - I.N.S. U Spec. Prog. 914050rsay, France

KEY WORDS: Scleractinian skeletons, Biomineralization,

Ultrastructure,

Biochemical characterization,

Phy-

logeny. PALABRAS

CLAVE:

Esqueletos de escleractinios, Biomineralizaci6n, quimica, Filogenia.

Ultraestructura,

Caracterizaci6n

bio-

In Scleractinian coral taxonomy, major phylogenetic lineages are based on hypothetized relationships among septal microstructures, and various indications suggest that the widely used scheme proposed by WELLS (1956) have to be reexamined. The last decade of studies of biomineralization process have demostrated the leading role that specific macromolecules play during the growth of calcareous biocrystals. These results make possible a new microstructural approach, based on organo-mineral relationships examined at the micronic scale. In addition, biochemical informations obtained from extracted and purified macromolecules can be studied by multivariate analysis, resulting in diagrams showing the biochemical distances between species that can be compared to RNA/DNA based data. Thus, biochemically-based data provide us with evolutionary-related criteria by which phylogenetic distances can be asessed between the skeletal structures.

En corales escleractinios, las Ifneas mayores filogeneticas se basan en relaciones hipoteticas entre microestructuras septales, y varios indicios sugieren ahora que el esquema propuesto por WELLS (1956), muy ampliamente usado, debe ser revisado. En la ultima decada, los resultados de los estudios de procesos de biomineralizacion en organismos calcareos, y mas en concreto, la intervencion de macromoleculas organicas en la produccion de biocristales, ha sido aplicada a los esqueletos de corales. Ello hace posible una nueva aproximacion microestructural, basada en relaciones organominerales examinadas a escala microscopica. Ademas, la informacion bioquimica obtenida de la extraccion y purificacion de macromoleculas puede ser estudiada por analisis multivariante, que se reflejan en diagramas que pueden ser comparados con bases de datos de RNAIDNA. De esta forma, los datos bioqufmicos nos aportan criterios evolutivos que permiten calcular distancias filogeneticas entre las estructuras esqueleticas a partir de las cuales comienza el crecimiento de los corales. Esto permite desarrollar un dialogo con taxonomos de "tejidos blandos"

1.

MICROSTRUCTURE AND SCLERACTINIAN PHYLOGENY: PAST AND PRESENT

Sc1eractinian taxonomy is based on a very restricted number of concepts. It is generally agreed that septal microstructures provide us with the most important characters for defining phylogenetic lineages among families. The most formal statement of the major phylogenetic value of microstuctures

was made by WEl1..S(1956: F 339), and his relationship-tree of Sc1eractinian families (1956: F 363, fig 259), wich is without a doubt the most familiar picture to people dealing with post-Paleozoic corals, is based on an additional diagram that exposes Wells's conception of relationships among the sc1eractinian septal structures (WELLS, 1956: fig. 237). Since the publication of these forty-year-old pictures, the microstruc-

J.-P. CUIF. Y.

130

tural analyses that have been carried out have introduced considerable changes in our taxonomy view of Scleractinian.

l.l.

The discrepancies among recent - b as e d p ropo s al s

More recently, RoNrswrcz & MoRYcown (1989) recognized "seven microstructural groups" in their study of Triassic/Liassic Scleractinia, and subsequently, in a general study of Scleractinian evolution (RoNrnwrcz & Monvcowe,, 1993), they have suggested only four major "microstructurally

mi c ro s tructure

In the two more recent attempts

DAUPHIN and P GAUTRET

to

of Scleractinian classification, by Cuevnlrtn (1987), and

present a complete picture

RoNrswrcz & Monvcow,q. (1993). a prominent role is still given to septal structures in

defining the major subdivisions (Fig. l). However, in the first sentence of the chapter "Systematic", CHsvRt-tsR (in Grassé, 1987: 679) pointed out that unsolvable questions

defined groups including the previsionsly defined seven ones".

Comparing more precisely these two taxonomic proposals, we can observe that a given family can be placed in different microstructure-based groups. For instance, the suborder Distichophyllina is considered as "non trabecular" (CHrvelmn) but "microtrabecular" by Roniewicz and Morycowa.

Chevalier & Beauvais, 1987 -- - lno medio-sePtal Plan NO TRABECULE I

lmedio-sePtal Plan

TRABECULES Ino

fan-system

Relationships between septa and theca

Synapticules

I

Structure of trabecules

l'an-system

Roniewicz & Morycotua, 1993 FASCTCULAR MI NITR ABECULAR

TyPe of sePtal THICKTRABECULAR ornamentation

Arrangement

oftrabecules

Fig. 1.-Basis of major taxonomic subdivisions in two recent global classifications of Scleractinia. Septal microstructure remains the"first order criterios, but introduction of new microstructrural terms introduce unsolvable. los Scleractinia. La -Base de las subclivisiones taxon6micas mayores en dos recientes clasificaciones globales de microestructura septal sigue como criterio de primer orden, pero la introduccidn de nuevos términos microestructurales crea discrepancias insalvables.

have been introduced in the currently admitted views by occurrence of new data concerning Triassic corals. "Presently, there is r7o sotisÏactoryt Scleractinian systematic, particularly for fossil J'orms, and especially for those of early mesozoic". As a consequence, in the GnnssÉ Treatise there is no equivalent to Wells's phylogenetic picture. Bol. R. Soc. Esp. Hist. Nat. (Sec. Geol.),92 (1-1), 1997.

The Caryophylliina are classified as "trabecular" by Chevalier whereas this group is recognized as "minitrabecular" by Roniewicz &

Morycowa.

Obviously, the "trabecular" concept,

and more generaly microstructural terminology, needs to be clarified. The reality of the

"microstructural paradox" is demonstrated by this correlation between an increasing

BIOMINERALIZATION FEATURES IN SCLERACTINIAN CORAL SKELETONS

amount of microstructural data. the recosnition of new microstructural categories ànd, at the same time, the decreasing value of

skeletal characteristics

to provide a

solid

basis for assessing different these proposals.

1.2. Emergence phylogenies

There

of

"non-microstructural"

is an other strons

urges palaeontologists to

131

To realize this obiective. the confused situation i n microstructural analvsis reo u i res a thorough ancl determined revision of the concepts on which current Scleractinian taxonomy is based.

With this in mind, we would like to emphasize here and give examples of basic biomineralization conceots. based on recent advances in the understanding of the biologi-

reason that

cal calcification process. (Fig. 2) This "bio-

improle the current

mineralization-based approach" of scleractin-

2

3

Fig. 2.-Schematic presentation of septal microstructure in Favia (1), Montastraea (2) ancl Montlivaltia (after Grr-l & Lepusrs, l97l), based on repartition of calcification centres. -Presentacidn esquemâtica dè la microestructura septal en Favia (.1), Montastraea (2) y Montlivalria (segrin Gu & Lapusre, 1971), basado en la distribucidn de los centros de nucleaci6n.

scleractinian systematic basis. Until now, biologists dealing with Scleractinian have been using the skeletal-based taxonomy we are so 1àmiliar with. Thus, as palaeontologists. we are used to believing thàt there is no gap between biological and palaeontological taxonomic frameworks. But this is turning to change. The advent of molecular techniques enables biolosists to

create their own major evolutive liÀeages. exclusively based on soft tissues, without any regard for skeletal characteristics. Clearly, the next decade will be marked by a fun-

damental re-examination

of

Scleractinian

sub-orders and families based on RNA/DNA sequencing. It is in the major interest of evo-

lutionary studies of Scleractinian evolution to avert the emergence of a systematic gap between palaeontologists and biologists. This can only be achieved by participating in this developing research, already being carried out in some laboratories.

ian skeletons should enable us to clarify the terminology problem we are now faced with, and at the same time, to create the needed link between "soft-tissue" and "skeletal" taxonomies.

2.

RpCgNT ADVANCES

IN

UNDERSTANDING

BIoMTNERAL FoRMATIoN: Nucr-earroN TEMPLATE AND CRYSTAL SHAPE CONTROL. THE DOUBLE ROLE OF ORGANIC MATRICES

The major results achieved in the field

of biomineral formation durins the last decade can be briefly summarized

- Formation and growth of mineralized skeletal units are the result of a continuous interplay between mineral lattices and macromolecular organic compounds; - These organic compounds, whose biochemical compositions are clearly taxonomylinked, remain within the skeletons, entrapped Bol. R. Soc. Esp. Hist. Nat. (Sec. Geol.), 92 (I-4), 1997.

in or around the crystalline units whose growth has been "organically controlled". Various models have been proposed to explain the basic features of this organo/mineral interplay, mostly based on the MITTERER CUNNINGHAM observation (1985) that the

ic (CUIF & GAUTRET,1995), with a varying glucidic/proteic ratio. (Fig. 3). The proteic part shows an extremely high concentration of aspartic acid which is believed to ensure the organo-mineral linkage (up to 50 -60 % of the whole amino-acid content).

300 200

•

Pay

e Mont

~

Cyph

Il1llI Poc 0] Psa

miS Leps

.. [ill

Lepa

~

0

o :it

III

...Iit"

~

IS

e: it" U ...I

II proteins

I

Fig. 3.- Total organic content (micro gig) and glucidic/proteic ratio in some scleractinian skeletons. -Contenido organico total (micro gig) y coeficiente glucfdico/proteico en algunos esqueletos de escleractfnios.

carboxyl group of organic molecules and the carbonate ion of the mineral lattice exhibit close steric similarities. As pointed out by SIKESet al. (opening conference of the VIIth Biomineralization Meeting, Oceanographic Museum of Monaco, Nov. 1993), full understanding at the atomic scale of the organo-mineral processes that lead to biomineral formation is not in sight. However, it is already possible to apply to more specific research other holistic and well established results of biomineralization studies, Two main general rules of biomineralization processes are: a) The active "bonding part" of organic molecules involved in the mineralization process, (i.e. the aspartic acid residues in the Weiner model - 1979) is borne by various macromolecular architectures, the compositions of which are clearly linked to the taxonomic position of studied groups. For instance, these macromolecules are mostly proteic in bivalve molluscs, whereas the glucidic part is preponderant in Cephalopods. In corals, the "main body" of these mineralizing macromolecules is glyco-prote-

b) Another interesting and significant result in biomineralization studies, is the ability of living organisms to simultaneously create different microstructural and mineralogical units by changing the biochemical composition of organic matrices that lead to biomineralization process. The high phylogenetic value of these microstructural sequences have been assessed for a longer time, for instance in molluscan bivalves (B0GGILD,1930). These two rules can help us,when studying coral skeletons, to create the coherent framework we need to overcome the current difficulties in Scleractinian classification. 3. FROMCRYSTALLINE GROWTH TOBIOMINERALIZATION PROCESSIN SCLERACTINIAN CORALS: A REEXAMINATION OF THE "SPHERULITIC CRYSTALLIZATION" CONCEPT Underlying the microstructural analysis of microstructural components in scleractinian skeletons, we find the concept of "spherulitic crystallization". This notion originates

in two papers by BRYAN& HILL (194la b), in which similarities between organic (i.e. biologic) and inorganic spherulitic arrangement of crystals are discussed. This idea was extremely successful, resulting in a calcification model which is still largely accepted, in spite of some questionable points. Following this model, any skeletal structure in Scleractinian skeleton is produced by a two-step process. - Firstly, "calcification centres" are formed, acting as starting points of crystal development. They generally appear as dark spots (WELLS, 1956: 336), their true nature still being under discussion. Concentrations of organic matrices are reported, but sometimes the dark spots are believed to be a purely optical phenomenon, due to the tiny size of the crystals. During recent years, converging observations support the occurrence of crystalline "seeds" in these calcification centers. These intacellular-produced low-Mg calcite crystals should be directly expelled from calicoblastic cells and act as an initial support for further crystal development (KINSHINGTON, 1980; GLADFELTER, 1983; CONSTANZ,1985, 1986; CONSTANZ& MEIKE, 1990). - From these calcification centre, fibrous crystalline units grow. In agreement with the original hypothesis of BRYANand HILL, it is currently accepted that fibre growth is "radial", fibres developping in a direction globally perpendicular to the calicoblastic cell layer. Successive growth steps are marked within crystalline units, leading to the formation of growth lines, as illustrated by Wells in dissepimental structures (1956: F. 336). SORAUF& JELL (1977), have studied traces of this "incremental growth" in septa of Desmophyllum cristagalli , as part of detailed studies carried out by SORAUFon relationships between crystalinitially orientations and coral architectures (1970, 1972). Generally speaking, the usual crystal-growth rules are still believed to be the main influence during this part of skeletal formation, as was initially proposed in the "spherulitic crystallization" concept. There is a variety evidence that this "spherulitic crystallization" concept, mostly based on purely crystallographic development of fibrous skeletal units, has to be included in a more biological view of skeletal development.

1. As a purely geometrical argument, the "radial growth" of skeleton crystals (i.e. perpendicular to basal ectoderm) is easy to refute, as brevionsly discussed by WISE (1972). Leaving aside "undifferenciated" microstructural organisations (if any), it is very easy to find convincing examples of strong obliquity between growth lines (the record of successive positions of the secretory cell layer), and crystal growth direction (PI. I, fig. 1). 2. "Incremental growth " of crystals, as illustrated by Sorauf and Jell, provides us with evidence of crystalline discontinuity within fibro-crystals. Pictures demonstrating "compositional" heterogeneity in fibres (i.e. heterogenety in repartition of organic and mineral phases) can be obtained by different etching techniques. The action of enzymatic solutions on organic-rich layers,. or conversely, resistence to acidic etching (PI. I, fig. 1), reveals organic-rich growth lines within the crystal-like structures. 3. It is also easy to demonstrate, using a simple preparative process, the presence of a two-phased organic matrix in coral skeletons. The biochemical data presented below (cf. V) have been carried out on organic products extracted from coral skeletons from Moorea Island (Polynesia), using a preparative process allowing "soluble" and "insoluble" matrices. 4. Futher evidence concerning both the presence and leading role of organic matrices in the formation of skeletal units is given by the action of dilute acid on a polished surface of an Acropora sp. skeleton (PI. I, figs. 2-4). The skeletal spinose processes (PI. I, fig. 2), more or less protruding at the outer surface (depending upon the species) are initiated by a linear concentration of "calcification centers". In a second step of skeletal formation, these skeletal axes are completely surrounded by a dense cortex of elongated sclerodermites, whose distal ends give a very typical scaly aspect to the external surface of most Acroporid corals (Pl.I, fig. 3). In polarized light, these sclerodermites demonstrate a very meandering growth direction. Obviously, the morphology of these very long and sinous skeletal units is not compatible with a radial direction of crystals from the calcification centres; factor must be taken into account to explain this morphology.

discrepancies between the resulting classifications clearly indicate that we have to create a much more differenciated microstructural terminology, in order for us to describe precisely the evolutionary changes that occurred in the secretory process of the basal ectoderm of Scleractinia. Obviously, the crystal-like corallian fibre is a material that has long been difficult to deal with, but he case-studies presented below, involving two "trabecular" species, show that precise analysis of biomineralization patterns within septal structures (which reflect changes in secretory activity of basal ectoderm during coral ontogenesis) can be carried out with simple preparation techniques based on the observation of organomineral relationships at the micronic scale. Results suggest that a biomineralizationbased approach to coral skeletons can help us to improve microstructural analysis. I. External and internal views of septal growth process in Favia stelligera (specimens from lagoon of Moorea island, Polynesia). An upper view - at medium enlargment of a septal growing edge (PI. I, fig. 5) clearly shows the numerous skeletal axes whose growth is strongly oblique with respect to the general growth orientation of the coral surface. The growing ends of axis show a microcrystalline organization (PI. I, fig. 6) and, on the septal growing edge itself (PI. I, fig. 7), some patches with similar microcrystal clusters demonstrate that axes are initiated in these points and diverge (immediately and considerably) from the general growth direc4. TRABECULAE:IS THISCENTRALCONCEPT tion of the septum. After enzymatic etching, a REALLYEFFECTIVEFOR DESCRIBINGTHE polished surface perpendicular to the growth direction clearly reveals a two-phase process DIVERSITYOF SEPTALSTRUCTURES? in septal development: The first phase consists of the producSorauf has written (1993: 65): "virtualtion of separate calcification centres, initiatly all post-Triassic scleractinians have a traed by the nucleating cells of the calicoblastic becular septal structure". Thus, the central ectoderm. In some places, these calcification role of the "trabecular concept" coul not be centres develop in lateral axes, which are more emphasized. At the same time, howevclearly visible in diverse sections - arrrows. er, this postulated uniformity among septal (PI. II, fig. I ). structures appears to be a fundamental cause The second secretory phase is completeof the current status of microstructure-based ly different from a microstructural standmain phylogenetic lineages. What is the point. Most of the septal thickness is built by systematie value of trabeculae if all Scleracundifferenciated sclerenchyme, in which tinia are trabecular? The newly proposed successive growth lines are clearly visible microstructural terms (CHEVALIER, 1987; (PI. II, fig.3). RONIEWICZ& MORICOWA,1993) represent an 2. Another recognized trabecular speattempt to increase the analytical value of cies, Montastrea curta, demonstrates a rather the trabecular concept. But, the observed

Therefore it is interesting to note the difference in reaction to diluted acid etching (formic acid I %-5 seconds) between the calcification centres of the skeletal axis of this Acropora and the lateral parts of the sinuous units that build the second growth step of the skeleton. Whereas the calcification centres are highly soluble (as usual), the marginal parts of the cortical units are more acidresistant than the infilling aragonite (PI. I, fig. 4). This lowering of aragonite solubility could be caused by a close association with the coating organic matrices that have directed crystal growth, allowing these highly differenciated sclerodermites to develop their typical narrow and sinuous shape. With regard to these converging evidences, it can be suggested that the "spherulitic crystallization process" could be replaced by a "biomineralization-based" model of the origin and growth of sclerae tinian skeletal units. In a purely schematic view of an "undifferenciated" septum, a line of specialized cells (nucleating cells) creates the growing edge of the septum by repetetively producing calcification centres. These are composed of organic matrices dedicated to nucleating the crystalline units and/or, according to various authors, the calcitic crystal "seeds". A further step in septal development includes an "incremental growth" process (SORAUF& JELL 1977), during which the size and shape of sclerodermites are permanently controlled by organic compounds.

similar growth process (specimen from Mururoa island, Polynesia - J.P. Chevalier coI. in the Museum National d'Histoire Naturelle, Paris). The "septal ornamentation" is made of multi spinose units, laid out on the septal edge with a regular orientation perpendicular to the septal plan (PI. II, figs. 3,4). The 8 to 12 spheroidal elements that form each of the septal "teeth" are distinct calcification centers, as is easily seen on a polished and enzyme-etched surface. They are not continuous vertical axes, but rather short cluster of calcification centers, with a very precise spatial repartition (PI. II, figs. 5, 6). In the second growth step, production of an undifferenciated fibrous tissue gives a smooth profile to septa (arrows in PI. II, fig. 6). But the specific organization of the first secretory phase, which seems to be highly significant in establishing the basic organization of septa, can be easily read within the completed coral skeleton. Clearly, the term "trabecular" does not adequately characterize these two different septal organizations, and is not a sufficiently discriminating basis, to create useful and reliable taxonomic framework. Conversely, the study of positions and spatial arrangements of calcification centers and the description of the related early biomineralization phases provide us with precise data on the organization of the "nucleating system" which is active at the growing edge of each septum. In addition, description of microstructural features of successive growth phases reveals the evolutionary changes in secretory activity of basal ectoderm. Thus, "undifferenciated microstructure" (if any), in which the nucleating line is straight and continuous, and "typical trabecular" in which clusters of calcification centres are associated in regularly spaced spots, are the two exterme ends of the structural evolution of septa. Between these two poles, the architectural patterns of septa should be described and compared on the basis of ectodermal secretory activity. But interest in a biomineralizationbased approach to coral skeletons is not restricted to the need for more accurate descriptions of septal microstructures. In addition, analyses and biochemicaLcomparisons of sclerodermite organic matrices provide us with an efficient means for assessing

the phylogenetic distances between the main categories of microstructures.

5.

BIOCHEMICAL COMPARISONS OF SKELETAL ORGANICMATRICES: BIOLOGICAL ASSESSEMENT OF PHYLOGENETIC DISTANCES AMONGSEPTALARCHITECTURES

Skeleton - associated organic compounds can be extracted and purified using a preparatory process that first separates water soluble and insoluble parts of organic components (CUIF & GAUTRET,1995). Primary data can be obtained on the bulk of these two phases by separated analysis of proteic and glucidic parts. After specific hydrolysis, reverse phase chromatography (for amino-acids) or ion-exchange chromatography (for monosaccharids) yields. estimates of the sugar and proteic contents in studied skeletons. If necessary, more detailed fractionation of soluble parts can be carried out by various chromatographic techniques (ex. HPLC gel-filtration or ion-exchange techniques). It has long been established that the biochemical compositions of mineralizing matrices are clearly linked to taxonomy at the highest level (Phyla and classes). We can reasonably assume that biochemical differences which are presently observable at lower taxonomic level (order and families) represents the evolutionary accumulation of minor changes that occurred in biochemical mechanisms of skeletal matrix production (CUIF et at., 1996). Thus, submitted to multivariate analyis, amino-acid and monosaccharid concentrations in skelettal matrices allow calculations to be made of biochemically-based distances among skeletal matrices, i.e. the skeletal structures from which they have been extracted. An trial attempt has been made here using specimens from Moorea Island (Polynesia) belonging to following species: Favia stelligera, Pavona varians, Montastrea curta, Leptoseris sp., Leptastrea sp., Psammocora protrudacea, Pocillopora verrucosa, Montipora sp., Cyphastrea sp. The analytical results are presented in Table I (sugars) and II (proteins). The resulting graphs in Figs. 4 and 5 exhibit significant similarities and differences. It is not surprizing that evolutary trees produced by proteic

and glucidic compounds are not exactly similar: accumulated changes in genetic code for glucidic and proteic part can be produced at different rates. However it is interesting to

traea. These convergent results, based on both detailed septal microstructure observations and independent analyses of glucidic and proteic components of skeletal matrices,

Table L-Monosaacharid concentration in soluble matrices of coral skeletons (measurements by ion ex-change chromatography -pulsed amperometric detection- after T.F.A. hydrolysis). -Concentraciones de monosacaridos en matrices solubles de esqueletos de corales (medido por cromatograffa de intercambio de iones -detecci6n amperometrica pulsada- tras hidr61isis T.F.A.). Leps Fuc DRib Rha Ara Gal Gle Man Frc Rib GaIN GleN

18,44 0 0,59 0,16 32,95 8,93 9,75 0,42 5,92 11,65 11,61

Pav

4,5 35,71 1,04 1,77 27,99 4,07 17,04 0,29 2,37 2,26 3,24

Cyph

20,76 3,7 2,61 4,26 27,83 5,47 12,61 2,32 9,73 4,91 8,12

Mont

30,63 0,64 0,88 2,43 34,87 6,09 8,62 2,41 6,48 4,55 4,83

Fav

Poc

31,24 0,82 1,22 5,84 36,87 4,41 10,38 2,11 1,63 3,57 4,03

24,03 0 2,45 0 7,32 11,6 10,31 2,65 9,64 20,54 14,11

LEPTOSERlS

LEPTOSERIS

CYPHASTREA

CYPHASTREA

LEPTASTREA MONSTASTREA FAVIA PSAMMOCOAA

27,14 0,00 2,55 6,13 27,34 5,54 11,59 2,31 2,82 6,03 10,87

LEPTASTREA

PSAMMOCOR~

FAVIA PAVONA

PAVONA

POCILLOPORA*

note that the biochemical-affinity method places a specimen of Psammocora sp. in close position to Montastraea. This result differs from the Wells proposal (the genus is included in the family Thamnasteriidae), and also from the Chevalier proposal (in which Psammocora is used to found a new Family: Psammocoridae). Interestingly, when using a sufficiently high magnification to loot at the septal growth edge of the used specimen of Psammocora, the position of calcification centres clearly shows affinities with Montas-

37,97 8,23 0,81 5,02 22,19 4,04 ,89 0 5,54 3,95 5,35

Lepa

MONSTASTREA

POCILLOPORA

Fig. 4.-Biochemically-based distances among the studied corals, using glucidic compositions of soluble matrices. Method: Ascending hierarchical classification - No predefined groups. -Distancias bioqufmicas entre los corales estudiados usando la composici6n glucfdica de las matrices solubles. Metodo: clasificaci6n jerarquica ascendente sin agrupaciones prefijadas.

Psa

Fig. 5.-Biochemically-based distances among the studied corals, using amino-acid concentrations in soluble matrices (same method as above). -Distancias bioqufmicas entre los corales estudiados usando concentraciones de aminoacidos en matrices solubles (metodo similar al anterior).

provide us with a example of the effectiveness of biochemical comparisons. These biochemically-based trees have a double interest for phylogenetics. l.-Presently, there is no relevant quantitative method for analysing the evolutionary relationships of topographical changes in the secretory activity of ectodermal cells which determine the various septal architectures. Improvement in the accuracy of descriptions will improve potential discrimination among them, but will not provide a rational assess-

carrying out biologically useful analyses of skeletal matrices of fossil corals. At the moment the response is clearly negative. Recent data indicates that post-mortem alteration of macromolecular compounds occurs very quickly. Even in the best preserved fos-

ment of their phylogenetic distances. Conversely, results of biochemical analyses are obtained from biologic compounds directly linked to skeleton formation process, and can be compared with rational and perfectible methods.

Table n.-Amino-acid concentrations in soluble matrices of coral skeletons. (PITC deriyatization after standard hydrolysis). -Concentraci6n de aminmicidos en matrices solubles de los esqueletos de corales (DeriYaci6n PITC tras hidr6lisis estandar). Fay

Poc

23,19 20,05

20,15 27,73

1,2 1,01 1,02

0,3 0 1,38

20,12 13,96 6,26 2,68 5,31

13,47 4,67 4,59 2,49 2,5

6,64 1,72

1,05 2,26

Pro Hypro

3,99 0

Tyr Phe

0 1,76

Leps

Pay

Cyph

Asp Glu

28,49 6,39

9,74 7,93

32,57 16,79

Lys His Arg

0,58 2,71 0,65

1,27 2,75 0,84

Gly Ala Val Leu Ile

18,05 9,73 3,02 1,76 1,51

Ser Thr

Mont

Psa

Lepa

12,34 12,44

22,84 21,86

33,47 17,36

0,96 0 0,93

0,69 3,01 2,45

0,33 1,28 0,35

0 0,71

15,4 5,9 4,7 1,56 4,2

13,98 6,54 4,49

20,83 9,03 3,73 2,12 3,34

16,45 4,72 2,67 1,14 1,85

17,1 3,55 3,45

3,55 2,98

5,83 3,68

5,49

4,49 6,47

4,61 4,36

5,2 4,27

15,97 0,66

4,91 0,85

9,48

6,42

0,61

1,93

7,3 1,19

6,99 0,2

5,56 0,49

2,5 3,4

0,09 1,91

0,39 1,9

0 1,56

0 1,23

0,52 1,21

0,2 1,24

2. The resulting "phylogenetic trees" represent the ideal basis for a dialogue with the "soft tissue taxonomists". Biochemical data extracted from the skeleton are acceptable for paleontolgists, because they are derived from by the solid structures we usually work with. Obviously, they are also directly linked to a major biological process, and the observable differences are related to changes in the genetic code. They are therofore part of the usual biological approach of the evolutive phenomena. 6.

CONCLUSION:FROMLIVINGTO FOSSIL CORALS

For palaeontologists: the question immediately arises about the possibility of

2,12 4,71 1,71

0,42

1,28 2,27

sils (aragonites of Cenozoic material from the Paris basin) and afortiori in Triassic outrops (Alakir Cay valley, South Anatolia), organic material can hardly be detected (CUIF et al., 1992; GAUTRET& MARIN, 1993). Biochemical characterizations indicate that the crystal-linked remaining matrices have preserved at least a part of their original properties. As a rule however, they do not provide a reliable basis for extensive biochemical study. The possible discovery of an exceptional specimen cannot compensate for the unrecoverable loss of information that begins even in the oldest skeletal parts of living species. This is another factor that contributes to the focusing of analytical efforts of paleontologists towards the living faunas. Not only should the impending gap between biologists and paleontologists be avoided through a

common effort to create the multicriterial taxonomic framework we need, but in addition, palaeontologists have the specific task of fitting fossil faunas into the progressively changing, biologically-based concept of scleractinian evolution.

The authors thank J. Sorauf for the revision of manuscript and suggestions. Received September 14, 1995 Accepted June 28, 1996

B000ILD, O. B. 1930. The shell structure of the Molluscs. D. Kg!. Danskee Vidensk.Gkr., Naturvidensk. og Mathern., 9(2-2): 231-326. BRYAN,W. H. & HILL, D. 1941. Spherulites and allied structures-part I.. Proc. Roy. Soc. Queensland, LII (6): 41-53. BRYAN,W. H. & HILL D. 1941. Spherulitic crystallisation as a mechanism of skeletal growth in the hexacorals. Proc. Roy. Soc. Queensland LII(9): 78-91.

CHEVALIER, J. P. 1987. Scleractiniaires. In: Grasse, Traite de Zoologie, J. P. Ed. III-3, Masson Ed. Paris CONSTANTZ,B. 1985. Taxon-specific diagenetic variation among scleractinian corals (Barbados, West Indies). - Proc. Fifth Intern. Coral Reef Congr. Tahiti, 2: 86-92, Tahiti 1986. Coral skeleton construction: a physicochemically dominated process. Palaios, 1: 152-158. CONSTANZ,B. & MEIKE, A. 1990. Calcite centers of calcification in Mussa angulosa (Scleractinia). - In : Origin Evolution and Modern Aspects of Biomineralization in Plants and Animals, R. E. CRICK Ed., 201-207, 7 fig., Plenum Press, New-York. CUIF, J. P., DENIS, A., GAUTRET,P., MARIN, E, MASTANDREA,A. & Russo, E 1992. Recherche sur I'alteration diagenetique des biomineralisations carbonatees: evolution de la phase organique intrasquelettique dans les polypiers aragonitiques de Madreporaires du Cenozo"ique (Bassin de Paris) et du Trias superieur (Dolomite et Turquie). C.R. Acad. Sci. Paris, 314: 1097-1102. CUIF, 1. P. & GAUTRET,P. 1995. Glucides et proteines de la matrice soluble des biocristaux de Scleractiniaires Acroporides C.R. Acad. Sc. Paris, 320: 273-278. CUlF, J. P., DAUPHIN,Y, DENIS A., GAUTRET,P. & MARIN E, 1996. The organo-mineral structure of coral-skeletons: a potential source of new criteria for Scleractinian taxonomy. Bull. Inst. Oceanog. Monaco. Spec. Issue 14(4): 359-367. GLADFELTER, E. H. 1983. Skeletal developement in Acropora cervicornis, II. diel patterns of calcium

PLATEI1LAMINAI Fig. I.-Example of obliquity between fibro-crystal growth direction and growth-lines. Growth lines has been evidenced by brief etching (5s) by dilute formic acid (I %) The very fine acidic-resistant lines that mark the successive growth steps suggest that each of these growth level is formed by a close association of mineral with organic phase, that prevents rapid dissolution of the aragonite fibres. (x 250). • -Oblicuidad entre la direcci6n del crecimiento del fibro-cristal y las Ifneas de crecimiento. Estas han sido resaltadas por una breve corrosi6n (5s) en acido f6rmico diluido (I %). Las Ifneas muy finas, resistentes al ataque, que marcan los pasos de crecimiento sucesivos, sugieren que cada uno de estos niveles de crecimiento esta formado por una intima asociaci6n entre el mineral y las fases organicas que evitan una rapida disoluci6n de las fibras de aragonito. (x250) Fig. 2.-Acropora sp. External view showing the spinose processes (x 25). -Acropora sp. Vista externa que muestra los procesos espinosos (x25). Fig. 3.-Acropora sp. Between the protruding spinose processes, the scaly aspect of skeletal tissue is visible (x 180). -Acropora sp. EI aspecto escamoso de! tejido esque!etico es visible entre los procesos espinosos salientes (x 180). Fig. 4.-Acropora sp. Each of this scaly unit is the distal end of an extremely long sclerodermite, of which external limits are clearly marked by acidic-etching. As in Fig. I , the resistance of sclerodermite lateral parts to acidic etching suggest a close association vith macromolecular organic compounds which playa leading role in sclerodermite growth. (x200). -Acropora sp .. Cada una de las escamas es el final distal de una esclerodermita extremadamente larga cuyos Ifmites externos estan claramente marc ados por el ataque con acido. Como en la Fig. I , la resistencia de las partes laterales de la esclerodermita al ataque con acido, sugiere una intima asociaci6n con los compuestos organicos macromoleculares que juegan un pape! principal en el crecimiento de la esclerodermita (x200). Fig. 5.-Favia stelligera: Upper view of two septa. The short axis, that are strongly diverging from septal plan, are clearly visible (x40). -Flavia stelligera: Vista superior de dos septos. Los ejes cortos claramente divergentes de los pianos septales son bien visibles (x40). Fig. 6.-Growing end of a lateral axis. Microcrystalline aspect of skelettal tissue (x 380). -Final del crecimiento de un eje lateral. Aspecto microcristalino del tejido esqueletico (x380). Fig. 7.-Growing edge of a septum. Calcification centers are distinctly visible (x260). -Borde de crecimiento de un septo. Los Cel}.lfos de calcificaci6n son claramente visibles (x260). Fig. 8.-Microcrystalline feature of crystals in a\::alcification center (x450). -Caracterfsticas microcristalinas de los cristales en un centro decalcificaci6n (x450).

carbonate accretion. Coral Reefs, 2: 91-100, 9 fig. New-York. GAUTRET,P. & MARIN, F. 1992. Evaluation of diagenesis in Scleractinian corals and calcified Demosponges by substitution indes measurements and intraskelettal organic matrix analysis. in Vlth. Int. Symp. on fossil Cnidaria, including Archaeocyatha and Porifera. Cour. Forsch-Inst. Senckenberg, 164:317 -327 GILL, G. A. & LAFusTE, J., 1971. Madreporaires simples du Dogger d' Afghanistan : etude sur les structuresde type "Montlivaltia". Mem. Soc. GeoI. Fr., N.S.,50-115: 1-40, PI. I-XII. KINCHINGTON,D. 1980. Localisation of intracellular calcium within the epidermis of a cool temperature coral. In Development and Cellular Biology of Coelenterates, P. Tardent, R. Tardent R., Eds. pp. 143-148, Elzevier, Amsterdam. MITTERER,R. H. & CUNNINGHAM,J. R., 1985. The interaction of natural organic matter with grain surfaces: implication for calcium carbonate precipitation. Soc. Econ. Paleont. Mineral., Spec. paper, 86:16-31. RONIEWICZ,E. & MORYCOWA,E. 1989. Triassic Scleractinia and the Triasic-Liassic boundary. - Mem. Ass. Australas. Paleontols. 8: 347-354.

1993. Evolution of Scleractinia in the light of microstructural data - Cour Forsch. Inst. Senckenberg, 164: 233-240. SIKES, C. S., WIERZBICKI,A. & FABRY,V. J. 1994. From atomic to global scales in Biomineralization. Bull. Instit. Oceanog. Monaco, 14(1): 1-47. SORAUF,J. E. & JELL, J. S., 1977. Structure and incremental growth in the ahermatypic coral Desmophyllum cristagalli from the North Atlantic Palaeontology, 20(1): 1-19. SORAUF, J. E. 1993. The coral skeleton: analogy and comparisons, Scleractinia, Rugosa and Tabulata Cour. Forsch.-Inst. Senckenberg, 164: 63-70. WEINER, S. 1979. Aspartic acid-rich proteins: Major components of the soluble matrix of Mollusks shells. Calcif. Tissue Int., 29: 163-167. WELLS, J. W., 1956. Part F, Coelenterata; Scleractinia In: Treatise on Invertebrate Paleontology, R. C. MOORE, Ed. Geol. Soc. of America and University of Kansas Press. WISE, S. W. JR. 1972. Observation of fasciculi on developmental surfaces of scleractinian coral, skeletons - Biomineralization, 6: 160-175.

PLATEII/LAMINA II Figs. 1, 2.-Polished and enzyme-etched surface in a septum of Favia stelligera. The two step process in septal formation of this species is clearly observable on these two pictures. The calcification centres (arrows) along the median line are the initial stages of the short and diverging lateral axis. (x150 and x330) - cf. also Fig. 2. -Superficie pulida y atacada con enzimas en un septo de Favia stelligera. Los dos procesos de formaci6n de los septos son observables en ambas fotograffas. Los centros de calcificaci6n (flechas) a 10 largo de la lfnea media son (os estados iniciales de los ejes laterales corlos y divergentes (x 150 y x330). Ver Figura 2. Figs. 3, 4.-Montastraea curta. - External view (xlO) and example of septal "ornamentation" (x 140). -Montastraea curta. Vista externa (xlO) y ejemplo de ornamentaci6n septal (xI40). Figs. 5, 6.-Montastraea curta. - Enzymatic etching of polished surface. The initial phase of the septal formation is followed by a second step. The initial structures are completely covered by undifferentiated fibrous tissue, but are clearly revealed by enzymatic etching. The calcification centres (arrows) and the limit between the two phases (arrows) are demonstrated. (x70 and x250) - Note that in Fig. 6 the dissolving action of enzyme solution produces finely excavated line corresponding to organic rich layers (at the difference of acidic etching). -Montastraea curta. Ataque enzimatico de la superficie pulida. La fase inicial de la formaci6n septal es seguida por un segundo paso. Las estructuras iniciales son cubiertas completamente por tejido fibroso indiferenciado, pero son resaltados claramente por ataque enzimatico. Se muestran los centros de calcificaci6n (flechas) y ellfmite entre las dos fases (flechas) (x70 y x250). N6tese que en la Figura 6 la acci6n disolvente de la soluci6n enzimatica produce lfneas finamente excavadas que corresponden a capas ricas en materia organica. Figs. 7, 8.-Psammocora protrudacea - External view (x8) and "ornemental structure" of septal growth edge (x 150). The similarity with Fig. 4 is obvious. Interestingly, the distance 'calculation based on biochemical analyses of skeletal matrices indicates indicates a close proximity with the Montastraea analyzed here (see text). This result suggests a good correlation between structural evolution of septa and biochemical evolution of skeletal matrices that lead the sclerodermite formation. -Psammocora protrudacea. Vista externa (x8) y "estructura ornamental" del borde de crecimiento septal (x 150). La similitud con la Figura 4 es obvia. Notablemente, el calculo de la distancia basado en analisis bioqufmicos de las matrices esqueleticas tambien indica...una gran proximidad con la Montastraea analizada aquf (ver texto). Este Y resultado sugiere una buena correlaci6n entre la evoluci6n estructural de los septos y la evoluci6n bioqufmica de las matrices esqueleticas que control an la formaci6n de esclerodermitas.