When Urey et al. (1951) raised the eventuality of biological influence on isotopic fractionation within a belemnite rostrum from the Isle of Skye, this hypothesis must have seemed surprising because it concerned crystallization of Ca-carbonate, which was then regarded as a physical process governed solely by thermodynamics. Nothing in the earlier crystallization experiments carried out by McCrea (1950) allowed for consideration of possible biological influences. Paradoxically though, if biological influence would then have seemed improbable at the level of isotopic fractionation, more than a century of mineralogical observations and chemical measurements had already definitely established that crystallization of the carbonate of shells and coral skeletons was precisely linked to taxonomy, and therefore biologically controlled. It had been shown long before (Necker de Saussure 1839; Rose 1858) that molluscs produce both calcite and aragonite in a mode of deposition that is specific to each family, and that in certain species these two polymorphs of calcium carbonate are present within the same shell. Just the opposite, while corallites of the Scleractinia are exclusively aragonitic (Sorby 1879), the axes and spicules of corals in the Octocorallia are calcitic. It was understood, though, that the genus Heliopora builds its skeleton of fibrous aragonite.

It was also known that variations in environmental conditions tolerated by each species, and differences in physicochemical conditions which they can support, do not seem to determine whether changes occur in the polymorph of calcium carbonate deposited.

During the last few decades, research on climate change and threats to biodiversity have drawn public attention to present-day modifications of the world environment. Thus the public is now receptive to the concept of a changing Earth, a vision familiar to paleontologists for roughly two centuries, as investigations into fossilized life-forms began a few decades before the nineteenth century. A method for establishing the chronological distribution of the fossil record was rapidly developed and continuously improved, thereby providing access to Earth history for the first time. From a very restricted number of research centers, convergent information resulted in a completely new view of the Earth that outlines progressive changes in geography, and reconstructs environmental and faunal modifications through time. Also, up through the first half of the twentieth century, improving our description of the fossil record was an essential activity. It was the golden age of great paleontological monographs, in which the compositions of fossil faunas were compared in minute detail.

In the middle part of the twentieth century new physical methods were developed that led to innovative uses of fossils, many of which were based on study of the chemical record preserved in biogenic calcium carbonate of shells, reflecting environments in which the fossilized organisms lived. The first application of such a new method is precisely known: the reconstruction of annual variations in water temperature in Jurassic seas (about 150 million years before present), as a result of the study of variations in oxygen isotope ratios.

Research on the formation and properties of mineralized structures produced by living organisms is characterized by the diverse origins, methods, and aims of scientific investigators. This has resulted primarily from the rapid development of analytical methods in the second half of the twentieth century. As has been emphasized above in preceding chapters, the contrast between shell structures was never thought greater than at the beginning of this period. At that time, structures were still described with the optical microscope in the same way as Bowerbank had, yet, at the same time, were analyzed with the latest developments in theoretical and applied physics. These provided new data on isotopic fractionation in the calcium carbonate of the same shells.

Here we should emphasize that underestimating optical studies is a serious error. The logic of interpretations concerning the formation and method of growth of calcareous biominerals was influenced by their crystalline appearance in microstructural units, as observed with the polarizing microscope. It is important to acknowledge the great accuracy and precision of descriptions carried out at the end of the nineteenth and early twentieth centuries, and acknowledge our debt to these early researchers for the development of many of the major concepts in the field of biomineralization. Bourne (1899), in a seminal paper dealing with the formation of skeleton in the Anthozoa, carefully described the crucial zone of contact between the calicoblastic ectoderm and the corallite skeleton.

In addition to molluscs and corals, two groups that contain highly diversified contributors to biogenic carbonate sediments in present-day shallow seawaters, many other organisms are also important Ca-carbonate producers. For some of them, such as the Echinodermata, their present role only poorly reflects their former importance during older geological periods. This chapter, by no means exhaustive, aims to present some of their diverse three-dimensional organizations, focusing on the fine structure of calcareous hard parts. The underlying objective is first, to provide some insight into recognition of the main categories of Ca-carbonate producers of carbonate-dominated sedimentary facies. Additionally, since knowledge of diagenetic change is an obvious prerequisite for obtaining reliable evaluation of characteristics in fossilized biogenic materials, knowledge of the initial state of these Ca-carbonate materials provides the reader with a basis for comparison.

Benthic algae

Very few algae have developed calcareous biomineralization, but they play an important geological role: only two families of the Chlorophyceae and two others among the Rhodophyceae. From a viewpoint of methods, this calcareous mineralization provides remarkable contrasts between the two groups, as follows:

1. (1) They are mineralogical opposites, as calcifying Chlorophyceae virtually always produce aragonite, while the majority of the mineralizing Rhodophyceae form calcite (although Peysnella and related types utilize aragonite);

2. (2) They also are structural opposites, as the Chlorophyceae produce carbonate external to their tissues, while the Rhodophyceae develop calcification at the level of cellular membranes;

3. […]

Chapters 1 and 2 have established that the microstructural units forming mollusc and coral skeletons exhibit very comparable growth patterns. Coral fibers and mollusc prisms grow by the repeated production of mineralized layers, the thickness of which is in the micrometer range. Clearly, the unit of biomineralization is not the often-described microstructural component that forms each shell layer (prismatic, foliaceous, fibrous, granular, etc.), but rather it is the mineralization cycle that produces the elementary growth layer for each of the different components of the shells or corallites.

From a functional point of view, numerous examples show that distinct mineralized domains are simultaneously produced (e.g., Fig. 1.5i). The growth layer is isochronous, thus, the cyclical mineralizing process is developed on the whole mineralizing surface during the same time period, whatever the differences (microstructural and/or mineralogical) between contiguous mineralizing areas. In the first part of this chapter (Sections 3.1 and 3.2), the structure of the mineralizing growth layer here was investigated in both corals and molluscs, taking advantage of instruments that provide resolutions in the nanometer range. The spatial distribution of organic and mineral components is thus described within the elemental growth layer itself, allowing a functioning model to be inferred. The great similarity revealed by this high-resolution imaging fully confirms the results of the two previous chapters.

However, in contrast to the uniformity which can result from this common growth mode, attention has long been drawn to the potential of skeletal structures for providing information regarding the evolutionary process both in corals and molluscs.

Any improvement in our knowledge of the mineral structures formed by living organisms contributes to a more accurate analysis of their fossil remains. Such a simple remark is important, because selection of the methods used for assessing the preservation of any fossil, as well as interpretation of numerical values resulting from measurements carried out on fossil material, depends heavily on concepts regarding its original state and mode of growth. In corals, for instance, the amount of confidence in the reliability of isotopic or chemical measurements has long been based on a simple X-ray diffraction diagram, owing to postulation of a purely mineral composition of these “physiochemically” crystallized materials.

As defined by Berner (1980) the term “diagenesis,” as applied to any sedimentary object, refers to “the sum total of processes that produce changes – mineralogical, chemical and physical – from the time of deposition.” Such an extensive definition (see also Bates and Jackson 1980) obviously includes fossilization, the term we use when sedimentary processes are modifying materials that have been formed by living organisms. From this standpoint, the methods by which the three major biomineralization mechanisms control the deposition of their mineralized structures allow us to assume that diagenesis of the resulting materials will follow very different and specific pathways. A major difference from chemically precipitated crystals is that biominerals exhibit very distinct structural parameters at the micrometer and submicrometer scales, even if the chemical compositions of their mineral parts do not greatly differ from purely chemical equivalents.

In contrast to the process of calcareous biocrystallization, for which different interpretations are still being debated in current literature, questions concerning the state of the mineral phase in bones and teeth on the one hand, and in siliceous structures on the other, are not particularly a matter of controversy at comparable scales. As in calcareous structures, the observation of thin sections in polarized light with the optical microscope has permitted establishing the crystal organization of phosphatic material in bones and teeth, as well as the amorphous state of the silica in sponge spicules (as well as in diatoms, unicellular algae, and the Radiolaria). Surprisingly, questions that do arise concerning the more precise characterization of these two very different mineral phases have comparable causes. In bones and teeth, crystals are so small that, until electron microscopy, and now atomic force microscopy, direct observation was impossible. Yet bone and tooth constituents are by far the most studied of all mineralized biological structures, essentially in medical or dental research. The situation is still more frustrating concerning the mechanisms leading to the striking paradox in all siliceous biomineralized objects. In every case, morphology is seen clearly to be controlled, right up to the highest resolution of instrumentation, although this was first demonstrated in the spectacular drawings of microscopic observations of radiolarians made by Haeckel in the nineteenth century, and since completely confirmed by scanning electron microscopy.

The result of the initial step that began simply by comparing two types of calcium-carbonate material deposited by biological mechanisms (generally accepted as distinct), this study has moved far afield. Although not yet finished, this research provides truly convergent data regarding the methods of skeletal development that have resulted in calcareous structures produced by organisms. Now it is possible to suggest a generalized model for this biomineralization. The model essentially covers structures produced by epithelial mineralization, but, as this is a broadly shared type of calcification, the model pertains to most invertebrate animals. Coccoliths, unicellular algae with mineralization occurring within Golgi-derived vesicles, are the most notable exception.

In contrast to classical analysis of shell structures that were and are based on descriptions of the sizes and shapes of various mineralized units forming different and distinct shell layers, the methods used here focus rather on the introduction of time being an essential element of microstructural analysis. Determining what is synchronous in diverse structures, and still more informative, imaging the mineralized layers that contain actual evidence of the activity of mineralizing organs, now leads to examination of these deposits from a fresh viewpoint, far removed from the traditional approach.

According to this method, the actual growth unit is clearly a mineralized layer and a significant part of this study has been dedicated to the description of the structure and composition of exactly what the biomineralization unit is.

At the time that Heinz Lowenstam began an investigation bearing on the different minerals that living organisms can produce, only a dozen or so biogenic minerals were known. Thirty years later, their number surpasses 60. Well before the appearance of the synthesis uniting the essential data that had been established during the 1980s (Lowenstam and Weiner 1989), Lowenstam had proposed (1981) a fundamental distinction concerning the mode of formation of biogenic minerals, and more particularly the precision of the controls on their deposition exerted by the producing organism. Lowenstam proposed then to distinguish the “matrix-mediated minerals,” characterized by formation very precisely controlled by the action of an organic component specifically produced, and those that, although equally produced by a living organism, are developed in a more autonomous way; they are only “biologically induced.” In these last, mineral elements can be developed according to methods and arrangements quite close to those that can be observed in purely chemical precipitated materials.

Among the calcified structures belonging to this category were placed the calcareous skeletons produced by corals. This opposition between molluscan shells, examples of “matrix-mediated minerals,” and coral skeletal carbonate has remained generally accepted and very recently has still been formulated in reference journals (Veis 2005). This is however, truly surprising. The precision of the biological control on the skeleton by the coral polyps had been well established at the end of the nineteenth century by M. Ogilvie (1895, 1896), who carried out a pioneering study of the specific arrangements of the fibrous aragonite units there.

The microscopic study of living organisms did not develop until the first half of the nineteenth century, in spite of earlier observation of “cells” by Robert Hooke (1665). It was only in 1831 that Robert Brown described a new structure in plant cells: the “nucleus.” In transferring microscopic observations of Mathias Schleiden on plant tissues (1838) to his own studies on animal tissues, Theodor Schwann (1839) made a major contribution by proposing the theory of universal cellular organization in living organisms. In this same year the Microscopical Society of London was founded, and among the founding members was James Scott Bowerbank.

In his paper, “On the structure of the shells of molluscan and conchyferous animals” (1844), Bowerbank showed that far from being simple aggregates of unorganized mineral particles, these shells are formed of sharply distinct units, each possessing a definite geometry whose regular arrangement demonstrates that their order is completely controlled by the producing organism. The rapid diffusion of this new type of analysis is attested to by the publication the following year of Carpenter's paper, “On the microscopic structure of shells” (1845, with a second part in 1847).