Dissolution
Selective dissolution of skeletal elements
The susceptibility of skeletal grains to dissolution is
directly related to their mineralogy, their ultrastructure,
and their morphology. The bioclasts composed by the
most unstable mineralogical phases, such as HMC and
aragonite, are obviously the first to be dissolved. However, in details or during the early stages of dissolution,
differences are shown between elements of similar mineralogy. This differential susceptibility to dissolution
depends on the content in trace elements, notably the ratio
Mg/Ca of high-magnesian calcitic bioclasts and on the
fine-scale skeletal ultrastructure. This has been well
documented for coral skeletons, ultrastructural patterns
of which being taxon dependent. Hence, some coral taxa
have revealed more sensitive to dissolution than others,
and dissolution has been shown to occur preferentially at
particular sites, such as trabecular axes and/or median septal plans (James, 1974; Gvirtzman and Friedman, 1977;
Constantz, 1986; Dullo, 1986; Perrin, 2004). Additionally,
the earliest steps of dissolution evidenced from still living
colonies have been demonstrated to occur in the calcification centers, which are more abundant at these particular
sites, just a few years after the skeleton being secreted
(Perrin and Cuif, 2001). This appears to be strongly linked
with the rapid decay of intraskeletal organic matrices more
abundant in the centers of calcification (Perrin and Smith,
2007a, b).
Dissolution of early cements
During the diagenetic evolution of reefs, early cements
may be affected by following diagenetic processes including dissolution. Therefore, selective dissolution is not
restricted to the biogenic components. Similarly, the mineralogy, morphology, and trace element composition of
the cement are important factors controlling its susceptibility to dissolution. Aragonite, HMC, and, even, calcian
dolomite cements are variably susceptible to dissolution
(Purser and Schroeder, 1986). In the recent reefs of
Mururoa and Ouvéa, dissolution of botryoidal aragonite
occurs at microstructural scale forming microrhomboedral voids visible under SEM (Aïssaoui, 1985,
1986). Various degrees of partial dissolution have been
described in the Pleistocene facies of Mururoa Atoll,
affecting the isopachous fringes of fibrous highmagnesian calcitic cements (Aïssaoui et al., 1986;
Aïssaoui, 1988). However, the irregular dissolution pattern observed or its preferential location at some levels
within the thickness of cement layer remain difficult to
explain and may be related to a differential original composition of crystals forming the cement (Aïssaoui, 1986).
Dissolution of dolomite
The dissolution of dolomite or dedolomitization leads to
the formation of typical secondary voids which may be
common in Holocene, Pleistocene, or older reef rocks.
These traces of dedolomitization require to be identified
from thin sections. They are characterized by geometrical
voids, a few tens of microns to several millimeters in size,
which do not evoke the typical shape of any known
bioclast, the smaller voids having often a rhomboedral
periphery as does the dolomite crystal. Relics of small
crystals of calcian dolomite have been already described
on the walls of such secondary voids, hence leaving no
doubt regarding their origin (Aïssaoui, 1986). The secondary voids produced by the dissolution of dolomite in reef
carbonates can represent a quite important proportion of
the total rock, reaching in some cases 40%.
As a consequence, the significance of these dissolution
voids is extremely important for the interpretation and
reconstruction of reef history and the understanding
of reef diagenesis, since this secondary porosity is the
only evidence for an earlier stage of dolomitization. As
a consequence, dolomitization events have to be characterized through petrographical observations of thin sections and not solely through mineralogical analyses such
as x-ray diffractometry, for example. Moreover, the identification of dedolomitization voids may help in the precise
reconstruction of the initial geometry of dolomitic bodies
(Purser and Aïssaoui, 1985; Aïssaoui, 1986).
Distribution of dissolution features in reefs
Dissolution in reefs is a ubiquitous process, which often is
only detectable at the microscopic or even at the ultramicroscopic scale. The process itself starts at a very early
stage, in some cases just after the formation of the carbonate features which is affected by dissolution (Perrin and
Cuif, 2001; Perrin, 2004).
Larger-scale dissolution features in reefs are known to
occur during sea-level low stands through the exposure
of the marine carbonates to the effects of meteoric waters,
this leading to the formation of karstic framework and cavities especially under prevalent humid tropical climatic
conditions. These macro- to mesoscale dissolution features are distributed at specific horizontal intervals in the
ancient reef carbonates and are characterized by open
voids, chalky facies, and associated with other meteoric
and karstic features (e.g., typical cements and infills).
Controls of dissolution
The dissolution process is by definition selective in carbonates and this selection is operated through two main
controls: the fine-scale ultrastructure and the mineralogy
including the trace element composition. The ultrastructure determines the size and arrangement of individual
crystals and hence, within a given structure, either biogenic or nonbiogenic, the distribution of crystal boundaries, discontinuities, and crystalline faces which are
preferential sites, highly favorable to the initiation of dissolution process. This has been largely evidenced in various groups of skeletal carbonates (Walter, 1985; Dullo,
1986; Flessa and Brown, 1993; Perrin, 2004). In the case
of biogenic elements, the intraskeletal organic material
(proteins) and its degradation during the early stages of
DIAGENESIS
315
Selective dissolution of skeletal elements
The susceptibility of skeletal grains to dissolution is
directly related to their mineralogy, their ultrastructure,
and their morphology. The bioclasts composed by the
most unstable mineralogical phases, such as HMC and
aragonite, are obviously the first to be dissolved. However, in details or during the early stages of dissolution,
differences are shown between elements of similar mineralogy. This differential susceptibility to dissolution
depends on the content in trace elements, notably the ratio
Mg/Ca of high-magnesian calcitic bioclasts and on the
fine-scale skeletal ultrastructure. This has been well
documented for coral skeletons, ultrastructural patterns
of which being taxon dependent. Hence, some coral taxa
have revealed more sensitive to dissolution than others,
and dissolution has been shown to occur preferentially at
particular sites, such as trabecular axes and/or median septal plans (James, 1974; Gvirtzman and Friedman, 1977;
Constantz, 1986; Dullo, 1986; Perrin, 2004). Additionally,
the earliest steps of dissolution evidenced from still living
colonies have been demonstrated to occur in the calcification centers, which are more abundant at these particular
sites, just a few years after the skeleton being secreted
(Perrin and Cuif, 2001). This appears to be strongly linked
with the rapid decay of intraskeletal organic matrices more
abundant in the centers of calcification (Perrin and Smith,
2007a, b).
Dissolution of early cements
During the diagenetic evolution of reefs, early cements
may be affected by following diagenetic processes including dissolution. Therefore, selective dissolution is not
restricted to the biogenic components. Similarly, the mineralogy, morphology, and trace element composition of
the cement are important factors controlling its susceptibility to dissolution. Aragonite, HMC, and, even, calcian
dolomite cements are variably susceptible to dissolution
(Purser and Schroeder, 1986). In the recent reefs of
Mururoa and Ouvéa, dissolution of botryoidal aragonite
occurs at microstructural scale forming microrhomboedral voids visible under SEM (Aïssaoui, 1985,
1986). Various degrees of partial dissolution have been
described in the Pleistocene facies of Mururoa Atoll,
affecting the isopachous fringes of fibrous highmagnesian calcitic cements (Aïssaoui et al., 1986;
Aïssaoui, 1988). However, the irregular dissolution pattern observed or its preferential location at some levels
within the thickness of cement layer remain difficult to
explain and may be related to a differential original composition of crystals forming the cement (Aïssaoui, 1986).
Dissolution of dolomite
The dissolution of dolomite or dedolomitization leads to
the formation of typical secondary voids which may be
common in Holocene, Pleistocene, or older reef rocks.
These traces of dedolomitization require to be identified
from thin sections. They are characterized by geometrical
voids, a few tens of microns to several millimeters in size,
which do not evoke the typical shape of any known
bioclast, the smaller voids having often a rhomboedral
periphery as does the dolomite crystal. Relics of small
crystals of calcian dolomite have been already described
on the walls of such secondary voids, hence leaving no
doubt regarding their origin (Aïssaoui, 1986). The secondary voids produced by the dissolution of dolomite in reef
carbonates can represent a quite important proportion of
the total rock, reaching in some cases 40%.
As a consequence, the significance of these dissolution
voids is extremely important for the interpretation and
reconstruction of reef history and the understanding
of reef diagenesis, since this secondary porosity is the
only evidence for an earlier stage of dolomitization. As
a consequence, dolomitization events have to be characterized through petrographical observations of thin sections and not solely through mineralogical analyses such
as x-ray diffractometry, for example. Moreover, the identification of dedolomitization voids may help in the precise
reconstruction of the initial geometry of dolomitic bodies
(Purser and Aïssaoui, 1985; Aïssaoui, 1986).
Distribution of dissolution features in reefs
Dissolution in reefs is a ubiquitous process, which often is
only detectable at the microscopic or even at the ultramicroscopic scale. The process itself starts at a very early
stage, in some cases just after the formation of the carbonate features which is affected by dissolution (Perrin and
Cuif, 2001; Perrin, 2004).
Larger-scale dissolution features in reefs are known to
occur during sea-level low stands through the exposure
of the marine carbonates to the effects of meteoric waters,
this leading to the formation of karstic framework and cavities especially under prevalent humid tropical climatic
conditions. These macro- to mesoscale dissolution features are distributed at specific horizontal intervals in the
ancient reef carbonates and are characterized by open
voids, chalky facies, and associated with other meteoric
and karstic features (e.g., typical cements and infills).
Controls of dissolution
The dissolution process is by definition selective in carbonates and this selection is operated through two main
controls: the fine-scale ultrastructure and the mineralogy
including the trace element composition. The ultrastructure determines the size and arrangement of individual
crystals and hence, within a given structure, either biogenic or nonbiogenic, the distribution of crystal boundaries, discontinuities, and crystalline faces which are
preferential sites, highly favorable to the initiation of dissolution process. This has been largely evidenced in various groups of skeletal carbonates (Walter, 1985; Dullo,
1986; Flessa and Brown, 1993; Perrin, 2004). In the case
of biogenic elements, the intraskeletal organic material
(proteins) and its degradation during the early stages of
DIAGENESIS
315