Bacterially-mediated authigenesis of clays in phosphate stromatolites
A. SÁNCHEZ-NAVAS1, A. MARTÍN-ALGARRA1 AND F. NIETO1, 2
1. Departamento de Mineralogía y Petrología, Instituto Andaluz de Ciencias de la Tierra (C.S.I.C), Facultad
de Ciencias, Universidad de Granada, 18071 Granada, Spain (E-mails: asnavas@goliat.ugr.es, and
fnieto@goliat.ugr.es)
2. Departamento de Estratigrafía y Palaeontología, Instituto Andaluz de Ciencias de la Tierra (C.S.I.C),
Facultad de Ciencias, Universidad de Granada, 18071 Granada, Spain (E-mail: agustin@goliat.urg.es)
ABSTRACT
Authigenic clays in close textural relation to carbonate fluorapatite within finely laminated phosphate
stromatolites of Upper Jurassic age have been studied using scanning electron microscopy (SEM),
transmission electron microscopy (TEM) and analytical electron microscopy (AEM). Stromatolite laminae
consist of hexagonal prisms of francolite (sizes ranging between 0·1 and 1 μm) that are surrounded by poorly
crystalline smectite and amorphous Fe–Si–Al oxyhydroxides. Microanalyses show that smectite is Fe rich,
with highly variable composition, particularly regarding Fe and Si contents. Smectite has significant
beidellitic, montmorillonitic and non-tronitic substitutions. Although the lack of fringe contrast in some areas
adjacent to the smectite packets with 1·0–1·3 nm spacing is due to differences in orientation of layers,
textural and analytical data clearly indicate the presence of Fe–Si–Al amorphous phases intimately
intergrown with smectite. The occurrence of poorly crystalline smectite and associated amorphous phases
within microbially precipitated stromatolite laminae, both as envelopes around, and as pore-fillings between
extremely small calcium phosphate crystals, demonstrates authigenic smectite growth from a precursor Fe–
Si–Al amorphous material. This material is formed in close association with a phosphate-rich precursor. The
textural and structural relations, the preservation of chemical precursors of glauconite such as nontronitic
montmorillonite, and the presence of Fe–Si–Al amorphous mineral phases, imply crystallization of the
observed crystalline phases from synsedimentary (bacterially precipitated) amorphous precursors during
early diagenesis in postoxic environments. Carbonate fluorapatite was the first phase to crystallize from the
primary gel; smectite and associated amorphous Fe–Si–Al oxyhydroxides were the residual material of the
crystallization process. The slow rate of transformation (at low temperatures) from Fe–Si–Al-rich gels to
smectite, explains the textural relations between the poorly crystalline phases and the phosphate crystals, as
well as the preservation of amorphous substances in relation to clays. Authigenic smectite represents the first
step in glauconitization.
INTRODUCTION
An authigenic origin of clays associated with phosphorites has been inferred by analytical and X-ray
diffraction (XRD) studies. The relatively high iron content (≈7 wt % Fe 2O3) of montmorillonite intercalated
between primary phosphate laminae, and the relative sharpness of (0 0 1) XRD peaks were considered by
Nathan and Soudry (1982) and Soudry (1992)) as evidence of an authigenic origin. The substitution of Fe3+
and K+ in smectite is particularly important because of its implications for the genesis of several minerals, in
particular for that of the glauconitic mineral family (glaucony), the end-members of which are glauconitic
smectite and glauconitic mica according to the nomenclature of Odin and Matter (1981)). Nontronite occurs
as a widespread authigenic mineral in recent marine sediments. It has been studied in more detail than any
other smectite (Güven, 1988) and it is also a precursor of glauconite (Giresse & Odin, 1973; Velde, 1976;
Odom, 1976, 1984; Bautier et al., 1989). According to O’Brien et al. (1990), glaucony currently forming
today, together with apatite, on the East Australian continental margin is an immature, low-potassium, ironrich smectite. Glauconite/smectite mixed-layer material has been considered to have a complete range of
smectite and glauconite layers, and to precede glauconite formation (Velde & Odin, 1975) analogous to the
way in which illite/smectite mixed-layered material evolves to illite (Thompson & Hower, 1975; Bautier et
al., 1989).
The texture and structure of glaucony pellets have been described by Amouric and Parron (1985) using
transmission electron microscopy (TEM). They established that crystallization of glauconite occurs via a
smectite-like phase with d(0 0 1) = 12·5 Å (1·25 nm) which was probably derived from a precursor gel. Using
TEM and analytical electron microscopy (AEM), Eggleton (1987) studied the texture, structure and
chemistry of Fe–Si–Al-oxyhydroxide gels in soils, in order to establish their precise role in the weathering of
silicate minerals and in the neoformation of clays. These non-crystalline Fe–Si–Al oxyhydroxides form
hollow-packed spheres with walls showing an incipient layer structure, which evolve to clay minerals
(smectite or kaolinite, according to the gel composition).
A close textural association between Fe-rich smectite-like or glauconite-smectite phases and Fe–Si–Al
amorphous phases and francolite crystals has been found by Martín-Algarra and Sánchez-Navas (1995)
within open marine phosphate stromatolites. In this paper, we present a detailed textural, structural and
compositional study of these clays and associated amorphous phases, using scanning electron microscopy
(SEM), TEM and AEM.
MATERIAL AND METHODS
The clays described in this study occur in close textural relation to carbonate fluorapatite (francolite) within
phosphate stromatolites (centimetric oncoids and crusts) sampled in the Almola Sierra (Serranía de Ronda,
SW Spain; Martín-Algarra, 1987; Martín-Algarra & Sánchez-Navas, 1995). These stromatolites occur in
Upper Jurassic (latest Oxfordian–Kimmeridgian) condensed fossiliferous limestones, which were deposited
on a pelagic swell. A preliminary textural and mineralogical study of phosphate stromatolites was carried out
by optical microscopy, SEM and XRD.
Samples selected for TEM study correspond to primary (microbially controlled) stromatolitic phosphate
laminae (see Martín-Algarra & Sánchez-Navas, 1995; for details). They were detached from thin sections,
ion-milled and carbon coated. Lattice fringe images, qualitative analyses in TEM nannoprobe mode (19-nm
electron probe diameter) and quantitative chemical analyses in scanning tunnelling electron microscope
(STEM) mode of the various zones observed in the microtexture were obtained with a Philips CM20
instrument operated at 200 kV. An objective aperture of 40 μm was used as a compromise between optimum
amplitude and phase contrast for images. Reflections with d-values >0·4 nm were therefore used to obtain
lattice fringes. Quantitative microanalyses were carried out using a 5-nm beam diameter and a scanning area
of 100 × 20 nm. Muscovite, albite, biotite, spessartine, olivine and titanite were used to obtain k-factors to
correct energy dispersion X-ray (EDX) data by the thin-film method of Lorimer and Cliff (1976). Errors for
analysed elements (two standard deviations), expressed in percentage of the atomic proportions, are 6 (Na), 3
(Mg), 2 (Al), 4 (K), 4 (Ca), 5 (Ti), 3 (Mn) and 3 (Fe). Instrumental conditions for spectra acquisition were
200 s of live time for all elements except for K, for which a time of 30 s was used because of volatilization
problems.
RESULTS
TEM and SEM images
Stromatolite lamination is well defined macro- and microscopically by alternation of pristine, brownish to
yellowish phosphatic laminae, and darker laminae containing trapped clastic particles of calcite, quartz, Feoxyhydroxides, terrigenous clays and feldspars (Fig. 1). Phyllosilicates occur within non-phosphate laminae
as detrital mica (Fig. 2A,B), but also inside phosphate laminae, closely associated with francolite crystals.
Carbon-rich particles (organic matter) are also common (see Fig. 11D in Martín-Algarra & Sánchez-Navas,
1995).
In SEM images (Fig. 3), stromatolitic phosphate laminae appear as closely packed, ovoid to ellipsoidal
objects, sometimes grouped in complex aggregates or sausage-like strings. TEM images (Fig. 4) show that
most phosphate grains are hexagonal prisms (sizes ranging between 0·1 and 1 m) surrounded by a poorly
crystalline material that confers the rounded morphology to the phosphate grains observed under SEM. TEM
data show that this poorly crystalline material is formed by clay minerals associated with non-crystalline
substances (Fig. 5). Compositional (see below), textural and structural features (selected area electron
diffraction: SAED) of these phyllosilicates correspond to those of smectite. These include subparallel
arrangement of packets 20–50 nm in thickness, abundant layer terminations, and wavy layers (the common
fringe spacing is 1 nm: Figs 5 and 6). Although orientations within and between packets of smectite layers
are variable, they tend to form stacks consisting of a subparallel array of packets (Fig. 5). Some areas which
lack fringe contrast and which are adjacent to smectite packets with 1–1·3 nm spacing, have been confirmed
to correspond (as noted by Peacor, 1992), to smectite packets with different orientations than those with
fringe contrast (central areas of Figs 6 and 7).
The existence of Fe–Si–Al-rich non-crystalline substances has been established based on the following data:
(1) lack of a diffraction pattern (SAED photographs were quickly taken before exposure to the beam for
other purposes in order to minimize damage); (2) absence of lattice fringes and of Moiré pattern; (3) no
change of amplitude contrast of images when sample is tilted; (4) mottled appearance when a marked dark
constrast is observed; (5) variable chemical compositions producing anomalous stoichiometries on smectite
analyses or compositions that do not correspond to any observed crystalline phase (see below). Noncrystalline areas with a mottled appearance and associated with smectite (Figs 6–9) sometimes form
aggregates of rounded structures (arrows in Fig. 9A) similar to the spherical structures shown by ferrihydrite
gels in Eggleton (1987; Fig. 2). non-crystalline phases appear both within (intrapackets: Fig. 5) and between
(interpackets: Fig. 8) smectite packets. The observed textures demonstrate that smectite and amorphous
substances are intergrown at diverse scales, an observation that points to a common origin.
Chemical composition of smectite and associated amorphous phases
Smectite formulae, obtained from chemical analyses of clays by AEM, are listed in Table 1. Because of the
large dimensions of the areas of analysis (100 × 20 nm) relative to the small crystal sizes of smectite packets
(5–20 nm normal to the layer), and to the intimate intergrowth of smectites and non-crystalline phases,
quantitative analysis of pure smectite or only of the associated non-crystalline phase, without contamination
by another phase, proved difficult. Two quantitative analyses (squares in Fig. 10A) correspond to an Fe-rich
amorphous substance, one of which is probably a ferrihydrite (compare with Fig. 1 of Eggleton, 1987), as
confirmed by textural observation in TEM images (see above: Fig. 9A,B). Those corresponding to smectite
have been confirmed according to crystal chemical constraints. Taking into account Güven's (1988) data and
the precision of our AEM results, analyses corresponding to smectite have been selected according to the
following criteria:
Si: 3∙5-4
∑(VI): 1∙95-2∙2 or 2∙8-3∙05
∑(XII): <0∙65
where ∑(VI) and ∑(XII) represent, respectively, the sum of octahedral and interlayer cations.
Some of the selected analyses show particularly high interlayer charge, which results from the summation of
interlayer cations multiplied by their respective charges (K + Na + 2Ca in this case), the overall range being
0·23 and 0·87. The larger interlayer charge values are normally associated with illite, but a significant
proportion of the high charges is due to the inclusion of unknown quantities of amorphous domains in the
analyses (because very small areas of such material were undetectable, despite care taken in analytical
procedures – see below). In addition, AEM analyses include very small volumes of material, and the large
range of charge may simply because of local heterogeneities which are typical of a range of low- to highcharge smectite, in comparison with averaged bulk analyses. Such heterogeneity in smectite has been
implied by AEM analyses of several authors (Bautier et al., 1992; Freed & Peacor, 1992; Nieto et al., 1994,
1996). Other analyses could be considered as intermediate in octahedral occupancy to ideal di- or
trioctahedral smectites, but these have not been accepted because such smectite has rarely been reported
previously, and such `ideal' compositions are always derived from bulk chemical analyses. Moreover,
experimental syntheses of smectite with intermediate compositions have produced only a physical mixture of
dioctahedral and trioctahedral smectites (Grauby et al., 1993; Yamada et al., 1991). Analyses not reproduced
in Table 1 were considered to correspond to mixtures of clay minerals and associated amorphous phases. In
several analyses (e.g. Analysis 1, Table 1) an excess of Ca was detected that cannot be incorporated in the
smectite structure, and that necessarily must correspond to the adjacent non-crystalline material.
Qualitative nanoprobe analyses of the amorphous areas indicate that they are Fe-rich (Fig. 11A,C,D), with
minor amounts of Si, Al, Ca, K, Mg, P, S, Cl, and Na, in decreasing order of abundance. This is in agreement
with the dark contrast, which implies relatively high atomic number elements in these Fe-rich areas.
Although most of the analyses from areas corresponding to a mixture of smectite and amorphous phases
project onto smectite fields (montmorillonite–nontronite) in the Si–Al–Fe triangular composition diagram,
two clusters are observed, one close to the Si apex and another close to the Fe apex (Fig. 10A). The presence
of Si-rich amorphous phases was also detected by qualitative nannoprobe analyses (Fig. 11A and B>) inside
zones where analyses of mixtures between smectite and non-crystalline phases were obtained (triangles
nearest to the Si apex in Fig. 10A), although their presence was difficult to detect in TEM images.
Microanalyses show that smectite has a highly variable composition, particularly regarding Si, Mg and Fe
contents (Table 1). Most of the analyses may be plotted on the Al2–AlFe–AlMg ternary diagram (Fig. 10B),
and may be designated as Fe-rich montmorillonite according to Güven (1988). They have both important
beidellitic (Si → AlIV) and montmorillonitic (AlVI → Mg) components that produce a particularly high
interlayer charge, because both substitutions are compensated by an increase of the total positive charge of
the interlayer (K + Na + 2Ca) (Fig. 12, Table 2). In the case of both substitution mechanisms, our data vary
over nearly all of the previously described range of compositions (Güven, 1988): AlIV from 0·1–0·5 atoms
per formula unit (hereafter atoms p.f.u.); Mg from 0·2–0·6 atoms p.f.u.. In addition, the nontronitic
component is also very important (Fig. 12, Table 2), with Fe ranging from 0·16–0·65 atoms p.f.u. (Table 1).
Principal component analysis has been performed to quantify these substitutions (Labotka, 1983), using the
covariance matrix of Si, AlVI, Fe, Mg, K + Na + 2Ca, and Σ(VI) (Table 3). This analysis considers the cations
and vacancies (□) as coordinate axes in n-dimensional space, and each smectite composition is considered to
be a point in this n-space. Principal component analysis represents a transformation to a new set of
coordinates that are linear combinations of the original cation coordinates. This new set of coordinates is
chosen so that the new matrix, formed from the covariances of smectite compositions expresed in terms of
the new coordinates, is diagonal. The diagonal elements of the new matrix are the eigenvalues of the original
covariance matrix. The new coordinate axes are the eigenvectors of this matrix, called principal components,
and represent the variation in the chemical composition of the smectite (Table 3). The sum of the eigenvalues
is identical to the total variance, and the relative contribution of each eigenvector to the total variance of the
data set is given by the associated eigenvalue. The first principal component has the largest associated
eigenvalue, the second principal component has the second largest eigenvalue, and so on. The first principal
component (eigenvector I in Table 3) explains ≈81% of the total variance: 0·731 Si + 2·040 AlVI + 0·069
Σ(VI) = 0·731 AlIV + 1·299 Fe3+ + 0·673 Mg + 1·582 (K + Na + 2Ca). This first component is, in turn,
decomposed as follows:
(a) Beidellitic component: 0·731 Si + 0·731 □XII= 0·731 AlIV + 0·731 (K + Na + 2Ca)
(b) Montmorillonitic component: 0·673AlVI + 0·673□XII= 0·673Mg + 0·673 (K + Na + 2 Ca)
(c) Nontronitic component: 1·299 AlVI= 1·299 Fe3+. Principal component analysis confirms the importance
of these three well-known substitutions in smectite, and especially that of nontronitic substitution in the clays
of this study.
DISCUSSION
Bacterially mediated authigenesis of smectite via precursor Fe-Si-Al gels
The studied clays and amorphous phases occur within stromatolites, laminated sedimentary structures
accreted at the sediment–water interface by bacterially mediated precipitation of minerals, mainly phosphates
(Martín-Algarra & Sánchez-Navas, 1995). Both clays and amorphous phases form the rind of the oval to
ellipsoidal objects that constitute the phosphate-rich stromatolitic laminae (Figs 3, 4). The size and the shape
of these objects, which have cores composed of francolite crystals, resemble internal moulds of coccoidal
and rod-shaped bacteria, which have been interpreted as fossil remnants of the stromatolite-building bacterial
community (compare with Lamboy, 1990, 1993; Krajewsky et al., 1994; see also Martín-Algarra & SánchezNavas, 1995; and references therein). Such intimate textural and structural relations between the francolite
crystals, smectite and the associated amorphous substances demonstrate their close genetic relations. This
implies that the chemical components of all these phases must have been concentrated and precipitated
within microbial mats, forming a common precursor for them (Fig. 13). This primary precursor was probably
an extracellular, mucilaginous, bacterial gel, rich in polysaccharides, where PO43−, Ca2+, CO32−, Fe, Al, Si,
and other elements were bound and concentrated from seawater, degrading organic matter and/or particulate
sediments trapped within the stromatolite structure.
Degradation and dehydration of a polysaccharide-rich primary bacterial gel, would have produced noncrystalline inorganic precursor gels of the observed mineral phases (francolite, smectite and Fe–Si–Al-rich
amorphous phases). The location of smectite and Fe–Si–Al-rich amorphous phases at grain boundaries and
filling pores between francolite crystals, demonstrate that francolite crystallized from the precursory
precipitate first, an Fe–Si–Al-rich precursor gel being a residue of this first crystallization process (Fig. 13).
Taking into account the poor crystallinity of the smectite, the common presence of amorphous phases
intimately associated with smectite, and the rather similar chemical composition of much smectite and Fe–
Si–Al amorphous substances, we infer that both phases come from a primary Fe–Si–Al-rich precursor gel.
The smectites should have formed authigenically by an incomplete process of crystallization of this primary
non-crystalline precursor. The observed amorphous phases can be considered either as residues of smectite
crystallization, or as partially dehydrated relicts of the primary gel having original compositions that made
the formation of smectite impossible. In this sense, although smectites have a wide range of compositions,
they are not exactly scavengers, and cannot incorporate any element present within the bacterially
precipitated precursor.
The ability of bacteria to bind, concentrate and cause crystallization of many different minerals (including
sulphides, phosphates, oxyhydroxides and carbonates) is well established (Lowenstam, 1981; Krumbein,
1983; Ehrlich, 1990). Cation exchange in cell walls and sorption of different metals, even from dilute
solutions and sometimes selectively, has also been demonstrated (Marquis et al., 1976; Mayers & Beveridge,
1989; Ferris et al., 1988). Bacterial surfaces are reactive sites that allow either direct nucleation of crystals or
interaction between metals and biofilms by complexation of metastable organo-mineral phases (Morel et al.,
1973; Deggens & Ittekkot, 1982; Jannasch, 1984; Ferris et al., 1986, 1987). Bacteria also play an important
role in the weathering of silicate minerals (Ferris et al., 1989a, b) and in nucleation of Fe-Si amorphous
compounds in geothermal environments (Ferris et al., 1986; Konhauser & Ferris, 1996; Jones & Renaut,
1997). Moreover, the extracellular formation of fine-grained, compositionally heterogeneous, non-crystalline
and poorly crystalline Fe–Si–Al, phosphate and metal precipitates around bacterial cells has been
demonstrated recently in the laboratory (Urrutia & Beveridge, 1993, 1994). These precipitates slowly evolve
to crystalline phases, among which short range-ordered aluminosilicates have been identified (Urrutia &
Beveridge, 1995). In conclusion, the inferred participation of bacteria in the extracellular concentration and
precipitation of phosphate-rich and Fe–Si–Al-rich substances fully agrees with geomicrobiological studies
undertaken both in the laboratory and in present-day natural environments.
Kinetics of the crystallization process
The formation of metastable Fe–Si–Al-rich and P-rich precursor gels instead of direct crystallization of
mineral phases is favoured by: (a) high degrees of supersaturation which create strong interactions among
involved ions, leading to irregular coordination complexes large enough in size to precipitate; (b) the
presence of organic compounds that favour stabilization of metastable inorganic amorphous precipitates (see
Termine et al., 1970; for amorphous calcium phosphate -ACP- stabilization); (c) the lower surface and strain
energy of bubbly or foamy structures, typical of non-crystalline materials (see Eggleton, 1987; for Fe–Si–Al
non-crystalline oxyhydroxides). Bacterial surface microenvironments are able to create rapid supersaturation
conditions by ionic complexation, that account for the development of a gelly like matrix formed of hydrous
(Fe, Al)-silicate precursors (Ferris et al., 1987;Urrutia & Beveridge, 1993, 1994, 1995).
Laboratory experiments demonstrate a rapid transformation of amorphous to crystalline phosphate phases
(Posner et al., 1984; Van Capellen & Berner, 1991; Krajewski et al., 1994) that occurs through a solutionmediated, solid–solid conversion (Boskey & Posner, 1973). Although the transformation of Fe–Si–Al-rich
precursor gels into smectites could be interpreted to occur by a similar mechanism, smectite crystallization
from a precursor Fe–Al–Si-rich gel could also have occurred via a bubble-wall mechanism, such as that
proposed by Eggleton (1987) for clay mineral formation in weathering profiles.
As demonstrated by the textural relations (outer position of Fe–Si–Al precipitates relative to francolite
crystals, and common preservation of Fe–Si–Al amorphous phases in close association with the smectites),
the crystallization rate of the smectite from its Fe–Si–Al-rich precursor was slower than that of the francolite
from its P-rich precursor (possibly ACP according to Martín-Algarra & Sánchez-Navas, 1995). This is
consistent with experimental evidence of high reaction rates for the transformation of ACP to hydroxyapatite
(Posner et al., 1984), and with slow reaction rates at low temperatures, that favour the persistence of
metastable phases (Berner, 1984).
Physicochemical conditions
The marked variability in composition from site to site of the amorphous phases and smectites located
among francolite crystals is interpreted as a consequence of the heterogenity of the precursor bacterial gel.
Growth of an array of francolite crystals subsequently produced isolated domains of gel with varied local
Fe–Si–Al-rich compositions. The variable compositions were retained because of the slow mobility of ions at
low temperatures, and lack of connectivity of gel domains. Where the original compositions were
appropriate, crystallization of smectite took place, its final composition being a consequence of the local
chemical environment. In this sense, Tardy et al. (1987) have shown that the stability field of smectites at
low temperatures depends on the chemical activities of the various elements. Equilibrium may be reached
only locally in a large variety of aqueous microsystems, which, therefore, produce an extreme diversity of
clay particle compositions within a given population, as in the studied smectites. In particular, a change from
the stability field of nontronite to that of beidellite and to amorphous silica is a consequence of increasing
H4SiO4 activity (Tardy et al., 1987; Fig. 2). Moreover, experimental studies have demonstrated that complete
solid solution does not occur between different smectite types; when the composition of the starting material
did not correspond to an `allowed' formula of smectite, a mixture of various types of smectites, ± silica, was
synthesized (Yamada et al., 1991). Regarding the physicochemical conditions of the sedimentary
environment, Tardy et al. (1987) also pointed out that montmorillonite forms in solutions of higher pH (7·4),
characterized by higher concentrations of cations and silica, than kaolinite (pH = 6·8). These conditions
could easily be reached inside, or immediately below, the microbial mat when crystallization of francolite
and of authigenic clays occurred (compare with pH data in Ferris et al., 1989a; and Urrutia & Beveridge,
1994). In spite of the great variety of compositions of smectites obtained, no trioctahedral smectite has been
found; such smectite requires a high activity of Mg, and consequently its presence is usually related to
volcanic influences, which are completely absent in the studied materials.
Crystallization of the studied smectites postdates that of francolite within empty bacterial cells. The
crystallization of Fe-rich montmorillonite was probably favoured by a decrease in oxygen fugacity due to the
decomposition of bacterial organic matter (evolution from oxic to postoxic environments; Fig. 13). This is
consistent with independent experimental work on low-temperature synthesis of iron-rich smectite from
silica and Fe-hydroxides, which is only possible under reducing conditions (Harder, 1978).
Some of the studied Fe–Si–Al-rich amorphous phases and Fe-rich smectite constitute chemical precursors of
glauconite. The studied phases therefore may represent the first step in the glauconitization process or, in
other words, the gradual passage between the Al-rich clay minerals and the Fe-rich glauconitic smectite,
invoked by Odin and Matter (1981). Although the formation of glauconite sensu stricto did not take place in
the studied samples, it is frequently associated with pelagic phosphate and Fe–Mn stromatolites of various
ages (Martín-Algarra & Vera, 1994; Vera & Martín-Algarra, 1994; Föllmi, 1989; and references therein) and
glauconitization is commonly associated with phosphatization (Bentor, 1980; Baturin, 1982; Jarvis, 1992;
and many others). The existence of confined microenvironments is necessary for concentration of the
chemical elements involved in the early stages of glauconite formation (Jiménez-Millán et al. 1998, and
references therein). An intense reworking, which is commonly detected in glauconite-bearing sediments, and
the repeated exhumation of the precursor authigenic clays, with resultant change from postoxic to oxic
environments and exposure to bottom waters rich in K, would probably favour the maturation of Fe-rich
smectites to glauconites. Although some evidence of reworking does exist in the studied phosphate
stromatolites (see Martín- Algarra & Sánchez-Navas, 1995 for details), this was probably not intense or
persistent enough to favour the Fe- and the K-enrichments necessary for the formation of glauconite s.s. from
the studied Fe-rich smectite. Indeed, just as illite is the common prograde product of aluminium-rich
smectite, glauconite is the product of iron-rich smectite (Buatier et al., 1989).
CONCLUSIONS
In the studied phosphate stromatolites the crystallization of Fe-rich smectite, which is a precursor to
glauconite, occurred from Fe–Si–Al gels during early diagenesis. These precursor gels, together with
associated P-rich gels, are of synsedimentary origin and were produced by stromatolite-forming bacteria.
Authigenesis of the studied clays and the physicochemical evolution of the stromatolitic microenvironment
during their formation can be summarized (Fig. 13) as follows:
1 formation of primary P–Si–Fe–Al-rich gels as extracellular mucilaginous bacterial precipitates under
oxidizing conditions, because these substances were rich in ferric compounds and PO43− was enriched in the
sediment surface by oxidative degradation of organic matter;
2 after death of the bacteria, carbonate fluorapatite crystallized first from the primary, organic-rich gel. The
Fe–Si–Al-rich products changed slowly and partially (when allowed by the local gel composition) into
poorly crystalline clays;
3 crystallization of Fe-rich montmorillonite occurred under reducing (postoxic) conditions, after oxygen
consumption within the microbial mat as a consequence of the oxidation of bacterial organic matter during
early diagenesis (possibly just below the living mat). Reactants were the Fe–Al–Si-rich gels remaining after
apatite formation. The authigenic smectite is inferred to constitute the first step of an incomplete
glauconitization process.
ACKNOWLEDGMENTS
The authors thank to Dr Karl Föllmi (University of Neuchatel), Dr Jacques Lucas (University of Strasbourg),
and especially to Dr Donald R. Peacor (University of Michigan), who critically read an early version of this
paper and greatly helped to improve the final version of the text. The deep and positive critical reviews by Dr
Ian Jarvis, Dr David Soudry, and Dr Julian E. Andrews are also gratefully acknowledged. Dr María del Mar
Abad (Centro de Instrumentación Científica, University of Granada) is also acknowledged for her help
during TEM study. This work was financed by the Research Projects PB96-1430 and PB96-1383 (CICYT,
Spain) and is a contribution of Research Group no. 4089 of the Junta de Andalucía.
REFERENCES
Amouric, M. & Parron, C. (1985) Structure and growth mechanism of glauconite as seen by high-resolution
transmission electron microscopy. Clays Clay Miner., 33 , 473-482.
Baturin, G.N. (1982) Phosphorites on the Sea Floor. Origin, Composition and Distribution. Devel. Sedimentol.,
33 , Elsevier, Amsterdam, 334 p.
Bentor, Y.K. (1980) Phosphorites: The unsolved problems. In: Marine Phosphorites – Geochemistry,
Occurrence, Genesis (Ed. by Y. K. Bentor), Spec. Publ. Soc. econ. Paleont. Miner., 29 , 3-18.
Berner, R.A. (1984) Kinetics of weathering and diagenesis. In: Kinetics of Geochemical Processes (Ed. by A.
C. Lasaga and R. J. Kirkpatrick), Miner. Soc. Am. Reviews in Mineralogy, 8 , 111-134.
Boskey, A.D. & Posner, A.S. (1973) Conversion of amorphous calcium phosphate to microcrystalline
hydroxyapatite. A pH dependent, solution-mediated, solid-solid conversion. Jour. Phys. Chem., 77 , 23132317.
Buatier, M., Honnorez, J., Ehret, G. (1989) Fe-smectite-glauconite transition in hydrothermal green clays from
the Galápagos spreading center. Clays Clay Miner., 37 , 532-541.
Buatier, M., Peacor, D., O'Neil, J. (1992) Smectite-illite transition in Barbados accretionary wedge sediments:
TEM and AEM evidence for dissolution/crystallization at low temperature. Clays Clay Miner., 40 , 65-80.
Deggens, E.T. & Ittekkot, V. (1982) In situ metal-staining of biological membranes in sediments. Nature, 298 ,
262-264.
Eggleton, R.A. (1987) Noncrystalline Fe–Si–Al-oxyhydroxides. Clays Clay Miner., 35 , 29-37.
Ehrlich, H.L. (1990) Geomicrobiology. Marcel Decker, New York.
Ferris, F.G., Beveridge, T.J., Fyfe, W.S. (1986) Iron-silica crystallite nucleation by bacteria in a geothermal
sediment. Nature, 320 , 609-611.
Ferris, F.G., Fyfe, W.S., Beveridge, T.J. (1987) Bacteria as nucleation sites for authigenic minerals in a metalcontaminated sediment. Chem. Geol., 63 , 225-232.
Ferris, F.G., Fyfe, W.S., Beveridge, T.J. (1988) Metallic ion binding by Bacillus subtilis. Implications for the
fossilization of microorganisms. Geology, 16 , 149-152.
Ferris, F.G., Schultze, S., Whitten, T.C., Fyfe, W.S., Beveridge, T.J. (1989a) Metal interactions with microbial
biofilms in acidic and neutral pH environments. Appl. Environ. Microbiol., 55 , 1249-1257.
Ferris, F.G., Shotyk, W., Fyfe, W.S. (1989b) Mineral formation and decomposition by microorganisms. In:
Metal Ions and Bacteria (Ed. by T. J. Beveridge and R. J. Doyle), pp. 413–441, Wiley, New York.
Föllmi, K.B. (1989) Evolution of mid-Cretaceous triad (platform carbonates, phosphatic sediments and pelagic
carbonates along the Northern Tethys margin. In: Lecture Notes in Earth Sciences (Ed. by S. Bhattacharji, G.
M. Friedman, H. J. Neugebauer, and A. Seilacher), p. 153.
Freed, R.L. & Peacor, D.R. (1992) Diagenesis and the formation of authigenic illite-rich crystals in Gulf Coast
shales: TEM study of clay separates. J. Sedim. Petrol., 62 , 220-234.
Giresse, P. & Odin, G.S. (1973) Nature minéralogique et origines des glauconites du plateau continental du
Gabon et du Congo. Sedimentology, 20 , 457-488.
Grauby, O., Petit, S., Decarreau, A., Baronnet, A. (1993) The beidellite-saponite series: an experimental
approach. Eur. J. Mineral., 5 , 623-635.
Güven, N. (1988) Smectites. In: Hydrous Phyllosilicates (Ed. by S. W. Bailey). Miner. Soc. Am. Reviews in
Mineralogy, 19 , 497-560.
Harder, H. (1978) Synthesis of iron layer silicate minerals under natural conditions. Clays Clay Miner., 26 , 6572.
Jannasch, H.W. (1984) Microbial processes at deep sea hydrothermal vents. In: Hydrothermal Processes at
Seafloor Spreading Centers (Ed. by P. A. Rona, K. Bostrom, L. Laubier and K. L. Smith Jr), pp. 677–709,
Plenum Publ. Corp., New York.
Jarvis, I. (1992) Sedimentology, geochemistry and origin of phosphatic chalks: the Upper Cretaceous deposits
of NW Europe. Sedimentology, 39 , 55-97.
Jiménez-Millán, J., Molina, J.M., Nieto, F., Nieto, L., Ruiz-Ortiz, P.A.(1998) Glauconite and phosphate peloids
in Mesozoic carbonate sediments (Eastern Subbetic Zone, Betic Cordilleras, SE Spain). Clay Miner., 32 (in
press).
Jones, B. & Renaut, R.W. (1997) Formation of silica oncoids around geysers and hot springs at El Tatio,
northern Chile. Sedimentology, 44 , 287-304.
Konhauser, K.O. & Ferris, F.G. (1996) Diversity of iron and silica precipitation by microbial mats in
hydrothermal waters, Iceland: implications for Precambrian iron formations. Geology, 24 , 323-326.
Krajewski, K.P., Van Capellen, P., Trichet, J. et al. (1994) Biological controls and apatite formation in
sedimentary environments. Eclogae geol. Helv., 87 /3, 701-745.
Krumbein, W.E (Ed.) (1983) Microbial Geochemistry. Blackwell, London.
Labotka, T.C. (1983) Analysis of the compositional variations of biotite in pelitic hornfelses from northeastern
Minnesota. Amer. Mineralog., 68 , 900-914.
Lamboy, M. (1990) Microbial mediation in phosphatogenesis: new data from the Cretaceous phosphatic chalks
of northern France. In: Phosphorite Research and Development (Ed. by A. J. G. Notholt and I. Jarvis), Geol.
Soc. Spec. Publ., 52 , 157-167.
Lamboy, M. (1993) Phosphatization of calcium carbonate in phosphorites: microstructure and importance.
Sedimentology, 40 , 53-62.
Lorimer, G.W. & Cliff, G. (1976) Analytical electron microscopy of minerals. In: Electron Microscopy in
Mineralogy (Ed. by H. R. Wenk), pp. 506–519. Springer, Berlin.
Lowenstam, H.A. (1981) Minerals formed by organisms. Science, 211 , 1126-1131.
Marquis, R.E., Mayzel, K., Carstensen, E.L. (1976) Cation exchange in cell walls of Gram-positive bacteria.
Can. J. Microbiol., 22 , 975-982.
Martín-Algarra, A. (1987) Evolución geológica alpina del contacto entre las Zonas Internas y las Zonas
Externas de la Cordillera Bética (sector central y occidental). thesis University of Granada, 1171 p.
Martín-Algarra, A. & Sánchez-Navas, A. (1995) Phosphate stromatolites from condensed cephalopod
limestones, Upper Jurassic, Southern Spain. Sedimentology, 42 , 893-919.
Martín-Algarra, A. & Vera, J.A. (1994) Mesozoic pelagic phosphate stromatolites from the Penibetic (Betic
Cordillera Southern Spain). In: Phanerozoic Stromatolites II (Ed. by J. Bertrand-Sarfati and C. Monty), pp.
345–391, Kluwer Acad. Publishers.
Mayers, J.T. & Beveridge, T.J. (1989) The sorption of metals to Bacillus subtilis from dilute solutions and
simulated Hamilton harbour (Lake Ontario) water. Can. J. Microbiol., 35 , 764-770.
Morel, F., McDuff, R.E., Morgan, J.J. (1973) Interactions and chemostasis in aquatic chemical systems. Role of
pH, pE, solubility and complexation. In: Trace Metals in Natural Waters (Ed. by P. C. Singer), pp. 157–200,
Ann Arbor Sci. Publ., Ann Arbor, Michigan.
Nathan, Y. & Soudry, D. (1982) Authigenic silicate minerals in phosphorites of the Negev, southern Israel. Clay
Miner., 17 , 249-254.
Nieto, F., Velilla, N., Peacor, D.R., Ortega-Huertas, M. (1994) Regional retrograde alteration of sub-greenschist
facies chlorite to smectite. Contrib. Mineral. Petrol., 115 , 243-252.
Nieto, F., Ortega-Huertas, M.N., Peacor, D.R., Arostegui, J. (1996) Evolution of illite/smectite from early
diagenesis through incipient metamorphism in sediments of the Basque-Cantabrian Basin. Clays Clay
Miner., 44 , 304-323.
Odin, G.S. & Matter, A. (1981) De glauconiarum origine. Sedimentology, 28 , 611-641.
Odom, I.E. (1976) Microstructure, mineralogy and chemistry of Cambrian glauconite pellets and glauconite,
central USA. Clays Clay Miner., 24 , 232-238.
Odom, I.E. (1984) Glauconite and celadonite minerals. Miner. Soc. Am. Rev. Mineral., 13 , 545-572.
O'Brien, G.W., Milnes, A.R., Veeh, H.H., et al. (1990) Sedimentation dynamics and redox iron-cycling:
controlling factors for the apatite–glauconite association on the East Australian continental margin. In:
Phosphorite Research andDevelopment (Ed. by A. J. G. Noltholt and I. Jarvis), Geol. Soc. London. Spec.
Publ., 52 , 61-86.
Peacor, D.R. (1992) Diagenesis and low-grade metamorphism of shales and slates. In: Minerals and Reactions
at Atomic Scale: Transmission Electron Microscopy (Ed. by P. R. Buseck), Miner. Soc. Am. Reviews in
Mineralogy, 27 , 335-380.
Posner, A.S., Blumenthal, N.C., Betts, F. (1984) Chemistry and structure of precipitated hydroxyapatites. In:
Phosphate Minerals (Ed. by J. O. Nriagu and P. B. Moore), pp. 330–350. New York, Springer-Verlag,
Soudry, D. (1992) Primary bedded phosphorites in the Campanian Mishash Formation, Negev, southern Israel.
Sed. Geol., 80 , 77-88.
Termine, J.D., Peckauskas, R.A., Posner, A.S. (1970) Calcium phosphate formation in vitro. II. Effects of
environment on amorphous-crystalline transformation. Arch. Biochem. Biophys., 140 , 318-325.
Tardy, Y., Duplay, J., Fritz, B. (1987) Stability fields of smectites and illites as a function of temperature and
chemical composition. In: Geochemistry and Mineral Formation in the Earth Surface (Ed. by R. RodríguezClemente and Y. Tardy) pp. 461–494. CSIC-CNRS, Madrid.
Thompson, G.R. & Hower, J. (1975) The mineralogy of glauconite. Clays Clay Miner., 23 , 289-300.
Urrutia, M.M. & Beveridge, T.J. (1993) Mechanisms of silicate binding to the bacterial cell wall in Bacillus
subtilis. J. Bacteriol., 175, 1936-45.
Urrutia, M.M. & Beveridge, T.J. (1994) Formation of fine-grained metal and silicate precipitates on a bacterial
surface (Bacillus subtilis). Chem. Geol., 116 , 261-280.
Urrutia, M.M. & Beveridge, T.J. (1995) Formation of short range ordered aluminosilicates in the presence of
bacterial surfaces (Bacillus subtilis) and organic ligands. Geoderma, 65 , 149-165.
Van Capellen, Ph & Berner, R.A. (1991) Fluorapatite crystal growth from modified seawater solutions.
Geochim. Cosmochim. Acta, 55 , 1219-1234.
Velde, B. (1976) The chemical evolution of glauconite pellets as seen by microprobe determinations. Miner.
Mag., 40 , 753-760.
Velde, B. & Odin, G.S. (1975) Further information related to the origin of glauconite. Clays Clay Miner., 23 ,
376-381.
Vera, J.A. & Martín-Algarra, A. (1994) Mesozoic stratigraphic breaks and pelagic stromatolites in the Betic
Cordillera, Southern Spain. In: Phanerozoic Stromatolites II (Ed. by J. Bertrand-Sarfati and C. Monty), pp.
319–344, Kluwer Acad. Publishers, Dordrecht.
Yamada, H., Nakazawa, H., Yoshioka, K., Fujita, T. (1991) Smectites in the Montmorillonite-Beidellite series.
Clay Miner., 26 , 359-369.