Chemical Geology 347 (2013) 20–33

Contents lists available at SciVerse ScienceDirect

Chemical Geology
journal homepage: www.elsevier.com/locate/chemgeo

Isotopic evidence for dolomite formation in soils

J.L. Díaz-Hernández a,⁎, A. Sánchez-Navas b, c, E. Reyes c
a
IFAPA Camino de Purchil, Área de Recursos Naturales, Consejería de Agricultura, Pesca y Medio Ambiente, Junta de Andalucía, Apartado 2027, 18080 Granada, Spain
b
Departamento de Mineralogía y Petrología, Universidad de Granada, Campus Fuentenueva, Granada, Spain
c
Instituto Andaluz de Ciencias de la Tierra IACT (CSIC–UGR), Avda. de las Palmeras, 18100 Armilla, Granada, Spain

a r t i c l e i n f o a b s t r a c t

Article history: Dolomite formation in soils constitute a particular challenge because of: 1) scant magnesium content in
Received 19 November 2012 continental environments as opposed to the marine medium, 2) the kinetic problem related to the incorporation
Received in revised form 25 March 2013 of magnesium into the carbonate, and 3) the unknown role of soil dolomites in the global carbon cycle.
Accepted 28 March 2013
Pedogenic dolomite formed at deeper soil levels (subsoil) before the development of petrocalcic horizon
Available online 6 April 2013
barriers was investigated in a semiarid region of SE Spain (Guadix–Baza basin). Mineralogical characterization,
Editor: U. Brand textural relationships and isotopic data concerning soil dolomite, together with the results of a precipitation
experiment, provided fuller knowledge of the processes and conditions governing neoformation of dolomite in
Keywords: these soils.
Carbon–oxygen isotopes In the study case, dolomite enrichment occurs beyond the limit of major biological activity, which coincides
Dolomite problem with the rooting depth of native perennial plants in the semiarid soils studied. Textural studies reveal the
Inorganic carbon corrosion of inherited dolomite crystals in the upper soil horizons and the formation of dolomite in depth
Pedogenic processes in relation to a clayey material, composed mainly of smectites. Stable isotope distribution in dolomites
Semiarid soils
throughout the profiles indicates a fractionation with depth. This is explained by the formation of dolomites
Soil carbon storage
after the dissolution of the pedogenic calcite. The calcite detected in the subsoil is interpreted here as a
precursor of the neoformed dolomites that transport the isotopic signal associated with biological activity
of soils to deeper layers. Dolomite formation appears to be favoured by the presence of clay minerals in
the precipitation media. Clays retain water during evapotranspiration stages, which drastically change the
transport properties of the media and promote the incorporation of Mg into the structure of the neoformed
Ca,Mg-carbonate. As confirmed by laboratory experiments, diffusion-controlled crystal-growth processes
lead to the formation a precursory “protodolomite” with disordered Ca,Mg distribution from a fluid locally
supersaturated in dolomite.
© 2013 Elsevier B.V. All rights reserved.

1. Introduction dissolution–precipitation in soils. In fact, pedogenesis begins with

The genesis of calcic and petrocalcic horizons in soils occurs as
part of carbonate dissolution–precipitation cycles (Gile et al., 1966;
Salomons and Mook, 1976). Detrital limestone and other calcareous
materials undergoing decarbonation constitute a source for secondary
carbonate accumulation in soils. Biogenic and atmogenic CO2 are alter-
native sources for the carbon fixation in the soil under physico-chemical
conditions stabilizing the carbonate ion. The pedogenic environment is
strongly influenced by biological activity: root mats, in addition to
associated lichens and microorganisms, contribute directly to mineral
Abbreviations: a.s.l., above sea level; AEM, analytical electron microscopy; apfu,
atoms per formula unit; BSE, back scattered electrons; DIC, dissolved inorganic carbon;
EDX, energy-dispersion X-ray; FESEM, field-emission scanning electron microscope;
MCL, Meteoric Calcite Line; MDL, Meteoric Dolomite Line; OA, oriented aggregated;
TEM, transmission electron microscopy; XRD, X-ray diffraction.
⁎ Corresponding author.
E-mail address: josel.diaz@juntadeandalucia.es (J.L. Díaz-Hernández).
intense biological activity in the weathering zone (Wright and Tucker,
1991; Wright, 1992; Cerling and Quade, 1993). Carbonic acid produced
in association with biological activity (mainly roots) is the main acid
involved in the dissolution of calcium carbonate. The leaching depth
of the dissolved carbonate augments with increasing mean annual
precipitation and temperature (Arkley, 1963; Stevenson et al., 2005).
Carbonate precipitation at the deeper layers results largely from the
declining PCO2 in the soil and from the higher concentrations of solutes
in the soil solution due to evapotranspiration. In this sense, the carbon-
ation process is favoured in arid–semiarid regimes. Large amounts of
soil inorganic carbon, stored in the form of calcium carbonate, have
been reported in cultivated mollisols under moderately cold tempera-
tures (5–5.5 °C), and with a mean annual precipitation between 581
and 397 mm (Mikhailova and Post, 2006; Mikhailova et al., 2009),
suggesting an increase in the carbonation process with agricultural
practices.
Isotopic relationships found in pedogenic carbonates in arid–semiarid
ecosystems indicate that carbonate dissolution–precipitation cycles in a

0009-2541/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.chemgeo.2013.03.018
 J.L. Díaz-Hernández et al. / Chemical Geology 347 (2013) 20–33
pedogenic environment provoke changes in the isotopic composition of
the precipitated carbonate, by inducing significant inputs of biogenic
carbon (Cerling, 1984). Biogenic carbon carries a unique isotopic signa-
ture, and the biological fractionation processes deplete 13C to negative
values (Salomons et al., 1977; Magaritz and Amiel, 1980; Magaritz
et al., 1981). Therefore, the pedogenic carbonate will usually have more
negative 13C than detrital carbonates.
Modern dolomites are formed only in certain environments such as
meteoric, marine, hypersaline, subsurface brines, and hydrothermal
(Hardie, 1987). The precipitation of dolomite and dolomitization of
previous Ca-rich carbonates are among the least-understood problems
encountered in the geochemistry of carbonates: e.g. surface seawater
is strongly oversaturated in dolomite, but little evidence is available
on the widespread dolomite precipitation in modern, open-marine
sediments. To explain the “dolomite problem”, models have emerged
ascribing microbial activity to the mediation of dolomite precipitation.
Various authors have focused their attention on the role played by the
bio-mediation that leads to CO2 enrichment and raises alkalinity,
and presumably drives dolomite formation in Mg-rich environments
(e.g. Wright, 1999). In any case, the occurrence of dolomite in the
geochemical systems of the earth's surface is often strongly controlled
by reaction kinetics (Arvidson and Mackenzie, 1999; Deelman, 2005)
and/or mass-transport processes.
The formation of dolomite in soils is rarely reported, its origin being
related to parent-material inheritance, to atmospheric Mg2+ wet
deposition, and to the addition of dolomite contained in atmospheric
dust (Ghebre-Egziabhier and St. Arnaud, 1983; Ming, 2002; Goddard
et al., 2007; Díaz-Hernández et al., 2011). The occurrence of neoformed
dolomite in soils is usually restricted to saline environments (Shermann
et al., 1962; Kohut et al., 1995). However, the presence of high-Mg
water solutions in soils derived from weathering of serpentinites and
basaltic rocks may lead to the formation of authigenic dolomite in
non-saline environments (Podwojewski, 1995; Capo et al., 2000).
Although the formation of dolomite in soils has been discussed (Drees
and Wilding, 1987; Sobecki and Karathanasis, 1987; Bui et al., 1990;
Whipkey and Hayob, 2008), no mechanism explaining the precipitation
of dolomite in soils has been reported. The incorporation of magnesium
into the calcite structure to form high magnesian-calcites and proto-
dolomites under surface conditions is broadly described in the
literature (Hardie, 1987 and references therein; Deleuze and Brantley,
1997). Sedimentary and diagenetic dolomites probably precipitated as
protodolomites and their isotopic composition was controlled by the
protodolomite–water fractionation (Fritz and Smith, 1970). In any
case, the incorporation of magnesium to the rhombohedral carbonate
structure is not easy and appears to be controlled by kinetic factors
(Mucci and Morse, 1983; Putnis et al., 1995; Fernández-Díaz et al.,
1996).
2. Scope of the investigation
In this paper, we study the occurrence of pedogenic dolomite in
soils of an arid/semiarid region. Calcite is by far the most abundant
and best-studied pedogenic carbonate. On the other hand, dolomite
is not usually evaluated in soils and its role in the pedogenic processes
is scarcely known. Here, we describe the formation of dolomite at soil
depths below indurated layers delimitated by petrocalcic horizons or
caliches. Deeper layers (subsoil) in these types of soils are scarcely
studied because caliches are hard to excavate and their features at
these depths are irrelevant for soil taxonomy or for judging the soil
as a support for plants.
Several reasons underlie this approach: 1) Knowledge of the dis-
tribution of carbon in the deep-soil-layer system in semiarid areas is
essential to assess soil carbonates (calcite and dolomite) as a sink
for atmospheric carbon. However, the carbon distribution in deeper
layers (subsoil) has rarely been determined (Jobággy and Jackson,
2000; Díaz-Hernández et al., 2003). 2) Well-ordered dolomite, together
21
with protodolomites, has recently been documented in soils (Whipkey
and Hayob, 2008). 3) The soils of the study area are particularly rich in
inherited dolomites and, therefore, could be suitable for the develop-
ment of pedogenic dolomites in addition to calcites. 4) Periodic mois-
ture movement to depths below solum horizons serves as a regulating
factor in maintaining soluble Mg at levels favourable to the precipitation
of low-magnesium calcites (St. Arnaud and Herbillon, 1973), and occa-
sionally also leads to the formation of dolomites. 5) The few studies
available concerning pedogenic dolomites (St. Arnaud, 1979; Botha
and Hughes, 1992; Capo et al., 2000) point to the occurrence of pedo-
genic dolomites in depth.
The main objective of this research is to demonstrate the forma-
tion of dolomite in soils. For this, we firstly studied compositional,
mineralogical, and textural variations throughout the soil profiles,
finding that dolomite was dissolved in superficial horizons and
precipitated at greater depths than was the pedogenic calcite. Also,
we analysed the isotopes in calcite and dolomite of the soil samples.
This second stage revealed that the biogenic isotopic signature of
the pedogenic calcite was transferred to a neoformed dolomite
through the dissolution–precipitation cycles that affected the soil
carbonates. Finally, we discuss the physico-chemical constraints on
the dolomite formation in subsoil. According to the model proposed
for the precipitation of dolomite: 1) inherited dolomite constitutes
the main source for the Mg 2+; 2) hydrodynamics in the soil studied
controls moisture and solute transport properties in the subsoil; and
3) the incorporation of magnesium to the carbonate structure in
Mg-rich confined porous media to form dolomite is explained from
the standpoint of kinetics.
3. Material and methods
3.1. Geological setting and climate
The Guadix–Baza basin is located in the Betic cordillera (SE Spain,
Fig. 1A). The area studied was developed as an endorheic depression
from the Late Neogene to the Late Pleistocene (Vera, 1970; Peña,
1985; García-Aguilar and Martín, 2000; Gibert et al., 2007, among
others), and occupies about 3500 km 2. During this period, the basin
was filled with alluvial and lacustrine materials for which the compo-
sition varies depending on the nature of source materials. Those from
the Subbetic ranges (North border) are rich in limestone, while those
from the Betic ranges (South border) are rich in quartzite, micaschist,
and dolomite (IGME, 1973–1999). The Guadix sub-basin is comprised
of fluvial sediments (silts, sands and conglomerates) and the Baza
sub-basin is formed mainly by lacustrine deposits (limestones, marly
limestones, and gypsum). The rocky relief surrounding the Guadix–
Baza depression coincides partially with the Atlantic–Mediterranean
divide. The depression is characterized by having elevations ranging
from 1500 m a.s.l. (above sea level) on the edges to 650 m a.s.l. at the
Guadiana Menor River, the main drain of the area.
The capture of the internal drainage network of the depression by
the Guadiana Menor River (a tributary of the Guadalquivir River)
modelled the landscape into two main geomorphological groups
(Díaz-Hernández and Juliá, 2006):
1) Old surfaces (350,000–205,000 yr): S1, S2, S3, consisting of
dissected piedmont plains gently sloping towards the centre of the
basin (Fig. 1B), capped by caliche soils with thick petrocalcic hori-
zons (Petric Calcisols and Chromic Luvisols, FAO-UNESCO, 1988)
(Calciorthids, Paleorthids, Paleargids, Haploxeralfs, Rodoxeralfs and
Xerochrepts, following the Soil Taxonomy, Soil Survey Staff, 1999).
Dolomite enrichment observed under the petrocalcic horizon (see
below) suggests that the formation of dolomite began within this
broad time range.
2) The youngest surfaces (115,000–48,000 yr): S4 (badlands, with
Gypsic, Calcaric and Eutric Regosols, FAO-UNESCO, 1988) (Xerortents,
22 J.L. Díaz-Hernández et al. / Chemical Geology 347 (2013) 20–33

Fig. 1. A) Location of the study area showing the position of 81 profiles sampled. B) Cross section indicated in panel A showing the spatial relationship between the geomorphological
units of the area.

Soil Survey Staff, 1999) and S5 (alluvials, with Calcaric Fluvisols,
FAO-UNESCO, 1988) (Xerofluvents, Soil Survey Staff, 1999).
The climate of the basin is semiarid, hot in summer, with frequent
frosts in winter. The mean annual temperature is 15 °C. Mean
annual precipitation is about 250–300 mm in the central part and
about 400 mm in the higher fans located along the mountain fronts
(MAPA, 1989). Rainfall occurs mainly during the autumn–winter
months; in summer, rain is scarce, falling as infrequent torrential
storms.
3.2. Sampling
Soil profiles were studied at 81 sites in the Guadix–Baza basin
(Fig. 1A), where the depth and features of the caliches could be accu-
rately determined. Natural or artificial pits reached an average depth
of about 200 cm (in several cases down to 400 cm) and the sampling
sites were distributed as evenly as possible, in 10 × 10 km cells to
simulate a probabilistic sampling, with one or two sites selected per
cell. Data from the deepest layers were not considered here because
of the statistical results from the few data (N b 4) are of doubtful
 J.L. Díaz-Hernández et al. / Chemical Geology 347 (2013) 20–33
reliability. Samples were taken at intervals of 20 cm deep after
cleaning the sections to expose fresh material. From the 81 profiles
here studied, 43 correspond to rainfed cultivated areas, 33 to
Mediterranean bush, and 5 to Mediterranean forest.
3.3. Carbon soil analysis and particle-size measures
Organic and inorganic carbon were analysed in sieved samples
(b2 mm) using standard techniques (USDA-NRCS, 1996). Organic
carbon was determined by the Walkley and Black method, the CaCO3
equivalent was measured by a manometric method (Loeppert and
Suárez, 1996), and the calcite and dolomite contents were inferred from
the CaCO3 equivalent together with the peak-height ratios detected in
the X-ray diffraction (XRD) powder diagrams. Soil-particle-size analysis
was performed by the Robinson pipette method (USDA-NRCS, 1996). A
total of 892 samples were analysed.
3.4. XRD methods
The specimens were ground to particle sizes b20 μm and examined
using a Philips PW 1710/00 diffractometer, with a graphite monochro-
mator, automatic slit, Cu–Kα radiation and on-line computer connec-
tion. Data were collected for 0.4 s integration time in 0.05° 2θ steps at
40 kV and 40 mA, in a 2θ interval between 3 and 70°.
The fine fraction (b2 mm) of the sample at 220 cm depth was
prepared by dispersing the sample with an ultrasonic probe and
pipetting the suspension to obtain the b 2 μm fraction (fraction 1).
Carbonates were then removed after attack with diluted acetic acid.
This fraction was centrifuged at 3000 rpm, and a second suspension
was obtained from the supernatant (fraction 2). The supernatant
liquid was deposited onto a circular glass slide for the oriented clay-
mineral aggregates (OA). The occurrence of smectite in the sample
was investigated by treating the OAs with ethylene glycol. X-ray
diffraction diagrams of the OAs were recorded using a PANalytical X'Pert
Pro diffractometer (Cu–Kα radiation, 45 kV, 40 mA) equipped with an
X'Celerator solid-state lineal detector. The diffraction patterns were
made by a continuous scan between 3 and 50° 2θ, with a 0.01° 2θ step
size and 20 s time step.
The data were processed using the XPowder® programme in order
to determine the qualitative mineral composition and XRD map
(Martín-Ramos, 2004).
3.5. Isotopic analysis
The soil samples used in this study, selected from among the 81
profiles on the basis of an even distribution by geomorphic units,
were ground to b 200 mesh to be later treated with 100% phosphoric
acid for 12 h in a thermostatic bath at 25 °C (McCrea, 1950). The sam-
ples containing calcite and dolomite carbonates were also treated
according to the method of Al-Aasm et al. (1990). This method con-
sists essentially of three steps: 1) obtaining CO2 after a 2-h reaction
with phosphoric acid at 25 °C (labelled as calcite); 2) after of 24 h
at 25 °C elimination of CO2 from residual calcite and a small quantity
of dolomite, and 3) obtaining CO2 from the continuation of the reaction
at 50 °C for 24 h that was labelled as dolomite. The carbon dioxide
resulting from both methods was analysed using a Finnigan Mat 251
mass dual inlet mass spectrometer. The experimental error was less
than ±0.1‰ for δ 13C and δ 18O. The in-house standards Carrara and
EEZ-1 were previously verified against NBS-18 and NBS-19. All the
samples were compared to a reference carbon dioxide obtained from
a calcite standard prepared at the same time. The oxygen isotope ratios
for dolomite were calculated taking into account the fractionation factor
for acid decomposition at 50 °C, 1.01057 for dolomite (Rosenbaum and
Sheppard, 1986) and at 25 °C, 1.01044 for calcite (Kim and O'Neil,
1997).
23
3.6. Electron microscopy
Two samples corresponding to the fine fraction of soil material
located below the petrocalcic horizon at 120 and 220 cm depths
(profile 46) were selected on the basis of their dolomite content.
They were consolidated and prepared as polished sections and later
carbon coated for the textural and mineralogical study by analytical
scanning electron microscopy. Back-scattered electron (BSE) images
and energy-dispersion X-ray (EDX) spot analyses (200 nm of diameter)
were performed with a high-resolution field-emission scanning electron
microscope (FESEM) Auriga (Carl Zeiss) equipped with a LinK INCA 200
(Oxford Instruments) microanalysis system operated at an acceleration
voltage of 10–20 kV.
For transmission electron microscopy (TEM), the same b 2 μm frac-
tion of the mudstone samples studied by powder-XRD was embedded
in an epoxy resin, reoriented, and sectioned by diamond-knife ultrami-
crotome. Diverse thin-sections below 100 nm were deposited on a
copper grid and coated with carbon for the TEM study. Analytical elec-
tron microscopy (AEM) and TEM images were obtained with a Philips
CM20 instrument operated at 200 kV and equipped with an EDX
model EDAX. Analysis of smectite and amorphous material tightly
intergrown with clay particles was performed in STEM mode using a
beam 5 nm in diameter and a scanning area of 20 × 100 nm. For quan-
titative micro-analyses, EDX data were corrected by the thin-film meth-
od of Lorimer and Cliff (1976). The k-factors were determined using
muscovite, albite, biotite, spessartine, olivine, and titanite standards.
The quantitative analysis corresponding to the smectite packet was
selected according to the following crystal-chemistry constraints:
Si: 3.5–4 apfu (atoms per formula unit); sum of octahedral cations
(∑VI): 1.95–2.2 apfu; and K + Na + 2Ca (∑XII): b1 apfu.
3.7. Experimental precipitation of magnesian calcites
Solution growth of magnesian calcites was carried out by bulk
crystallisation (free-drift experiments) where two solutions of rela-
tively easily dissolved salts (0.1 M Na2CO3 and 50 mM CaCl2·6H2O–
50 mM MgCl2·6H2O) were mixed, and the resulting solution became
supersaturated with respect to the less soluble (Ca,Mg)CO3 substance.
In one of the experiments, a low-viscous medium was prepared by
adding 0.1% of agar to water, boiling the solution, and cooling to
25 °C. Precipitates were later isolated and studied by XRD using the
XPowder® programme.
4. Results
4.1. Compositional, mineralogical and particle-size variations in the soil
profiles
The distribution of organic and inorganic carbon with depth varied
considerably in the soils of the study region (Fig. 2A–C; see also Fig. 1
in Díaz-Hernández, 2010), reflecting the heterogeneity of the area,
in which several soil types are represented (Díaz-Hernández et al.,
2003). Profiles of organic and inorganic carbon (calcite and dolomite)
distribution with depth for alluvial soils, badlands, and piedmont units
are depicted in Fig. 2A–C. Organic carbon decreases with depth in all
soil types, and it is particularly marked in the case of piedmonts
(Fig. 2A). Petrocalcic horizon, represented by the enrichment in calcite
content in depth, is detected mostly in piedmonts. This horizon appears
between 10 cm and 90 cm depth, with 30–130 cm of thickness (Fig. 2B,
see also Díaz-Hernández et al., 2003). An increase of the dolomite
content with depth is also better appreciated in piedmont profiles, usu-
ally below the petrocalcic horizon (compare Fig. 2B with C). In the soils
studied, dolomite enrichment can occur at great depth (below 200 cm),
as is the case of the profile PC-43 of the Fig. 2C. This can be explained by
the presence of Mg-rich substrates, constituted by altered ophitic rocks
in the subsoil. The development of the petrocalcic horizons and the
24 J.L. Díaz-Hernández et al. / Chemical Geology 347 (2013) 20–33
 J.L. Díaz-Hernández et al. / Chemical Geology 347 (2013) 20–33 25

Table 1
Arithmetic mean percentage (mean in %) of the organic and inorganic (calcite and dolomite) carbon content and of the diverse particle-size fractions for the specified depths in the
studied profiles (number of samples, N, for each depth is also indicated).

Depth (cm) Organic carbon Inorganic carbon Particle-size (b2 mm) Particle-size (>2 mm)

Calcite Dolomite Clay Silt Sand Gravels

N Mean (%) Mean (%) Mean (%) Mean (%) N Mean (%) N

0 81 1.82 29.35 9.19 19.48 31.26 42.26 69 34.44 81

20 81 1.13 37.60 8.42 19.82 29.59 50.60 55 34.83 68
40 81 0.67 42.59 10.34 19.38 25.58 55.04 43 37.85 60
60 81 0.46 42.23 12.95 15.51 28.27 56.22 43 43.14 57
80 81 0.34 39.35 14.68 13.62 26.45 59.93 46 39.89 65
100 81 0.25 35.90 15.24 12.04 23.43 64.53 43 37.73 65
120 81 0.19 32.64 16.16 11.88 23.96 64.16 50 39.05 67
140 75 0.16 31.41 16.83 11.46 21.49 67.05 42 35.55 65
160 64 0.15 29.78 15.81 11.08 19.66 69.26 37 41.63 59
180 55 0.12 31.25 13.28 10.94 22.82 66.23 35 41.14 48
200 44 0.13 28.09 16.97 12.72 20.46 66.82 31 45.99 40
220 27 0.11 33.63 18.33 12.12 17.99 69.89 17 53.06 24
240 19 0.12 38.25 13.57 13.76 25.17 61.08 14 52.15 17
260 12 0.14 37.31 16.26 10.18 15.63 74.19 6 51.91 9
280 9 0.15 28.81 12.73 14.57 17.92 67.51 5 56.14 7
300 8 0.18 26.78 12.77 18.89 19.18 61.94 4 59.46 6
320 8 0.22 25.03 13.51 11.72 29.10 59.18 4 55.71 6
340 4 0.22 31.14 14.00 9.20 15.66 75.14 2 58.12 3
Total 892 546 747

occurrence of associated dolomite enrichment in depth are better ob-
served in the case of piedmonts.
A total of 81 profiles were analysed, 8 corresponding to alluvial,
11 to badlands, and 62 to piedmonts, in order to gain statistical in-
formation concerning carbon distribution in depth for the soils
studied. Table 1 and Fig. 2D show the depth distribution of soil or-
ganic carbon content (from organic matter) and calcite and dolo-
mite content (soil inorganic carbon), expressed as an average
percentage. The average organic carbon decreased rapidly with
depth, tending to near-zero values below 120 cm (N = 81 pro-
files), and continued to decline gradually to a depth of 220 cm
(N = 27 profiles). The calcite content initially increased with
depth, showing minor amounts near the surface and a maximum
between 40 and 60 cm depth (43% in the bulge). Below this
depth the calcite diminished gradually to 200 cm in depth (around
30%). A second bulge at a depth of 200–280 cm appeared from the
profiles of the units of type S1. Polycyclic profiles showing this sec-
ond maximum of calcite accumulation within this depth range
were found for S1 unit (not shown). These minor maxima
correspond to hard substrates or paleosols in those units, probably
caused by regional sedimentological changes, and are not consid-
ered in this work. The average depth distribution of dolomite
began with a low content (9% in the superficial horizons) which
progressively increased from 40 cm, reaching 17% at 140 cm
depth. However, the depth distribution of the dolomite and calcite
contents proved somewhat irregular below 180 cm due to the low
number of samples for these depths (Table 1). Average particle-size
distributions throughout the soil profiles studied show high con-
tents of sand and gravels with respect to clays and silts (Fig. 2E).
The XRD study evidenced that the major minerals of soil samples
were quartz, calcite, and dolomite, with minor amounts of clays, feld-
spars, and micas. For dolomite-rich samples, superstructure reflection
(015) (the main reflection of well-ordered dolomites; Goldsmith and
Graf, 1958; Reeder and Sheppard, 1984) was present. The depth
distribution of quartz, calcite, and dolomite was shown in the XRD
mapping of Fig. 3, corresponding to a depth sequence of diffractogram
patterns for profile 46 (piedmont). This image shows a decrease in
quartz with depth on the basis of the change in the intensity of the
(101) reflection, and an interruption of the signal coinciding with
the maximum of calcite enrichment (100 cm depth). By contrast,
the dolomite (104 peak) showed an inverse distribution to that of
quartz (in-depth enrichment), with the same in-depth interruption
and lack in superficial horizon. Calcite enrichment between 60 and
130 cm of depth, as indicated by the increase in the intensity of the
(104) peak, coincided with the development of a petrocalcic horizon
of 70 cm thick in this profile.
4.2. Isotopic data
The frequency distribution of δ 13C and δ 18O of the calcite in the
study samples are shown in Fig. 4A and B, respectively. These values
were consistently negative, and δ 13C ranged from −13 to −1‰ and
δ18O from −10 to −2‰, with a normal distribution and modes around
−7‰ in both cases. These negative values indicate a contribution of
isotopically light carbonate. The histograms for the δ13C and δ 18O of
the dolomite showed bimodal distributions (Fig. 4C and D): the two
modal values for δ 13C were −5 and −1‰, and −6 and −3‰ for
δ18O. The scatter diagram relating both δ values for the dolomite sam-
ples also showed these two different tendencies in the histograms,
and two clusters may be identified within the diagram (Fig. 4E). The
average isotopic composition of the inherited dolomites from the coarse
fraction of the studied soils is also represented in the bivariate plot (star
in Fig. 4E). It plotted close to the analysed dolomite cluster with less
negative and even positive δ values (cluster I in Fig. 4E). Samples of
deeper layers (>70 cm) defined the cluster with more negative δ13C
and δ18O values (cluster II in Fig. 4E) that corresponded to the smallest
left-hand maximum observed in the histograms (Fig. 4C and D).
However, none of single profiles studied here showed a clear isotopic
fractionation with the dolomite enrichment in depth.

Fig. 2. A–C) Vertical distribution of organic and inorganic (calcite and dolomite) carbon in alluvial soils (PA, Fluvisols), badlands (PB, Regosols) and piedmonts (PC, Calcisols).
Geomorphological units are also indicated in brackets. Arrows in panel C mark dolomite enrichment in depth for the diverse type of soils. D) Depth-wise (y-axis) variation of
the organic carbon, calcite, and dolomite (inorganic carbon) content, expressed as an average percentage, in the study profiles of Guadix–Baza basin, drawn from the data of
Table 1. E) Distribution with depth of diverse granulometric fractions of soils; only mean values calculated from the studied profiles have been represented (gravel + fraction
b2 mm = 100% and contents of the fractions b2 mm (clay + silt + sand) are, in turn, recalculated to 100%). Panels D–E show the limits of the layers referred to in the text
(soil, petrocalcic, and subsoil). Superficial horizon (epipedon) and endopedon in the subsuperficial horizon are also indicated in these figures.
26 J.L. Díaz-Hernández et al. / Chemical Geology 347 (2013) 20–33

Fig. 3. False colour map drawn from XRD analyses of powdered samples (profile 46, piedmont S2). The y-axis represents the depth of the samples at intervals of 20 cm, with soil
surface on the top. Colour corresponds to the relative intensities of XRD peaks (26–32° 2θ interval only is represented). The approximate limits of petrocalcic horizon are indicated.

4.3. SEM and TEM study
Electron microscopy confirmed the XRD observations of Fig. 3
regarding dolomite dissolution and calcite precipitation in relation
to the formation of petrocalcic horizon. BSE images of Fig. 5A and B
correspond to a sample below the petrocalcic horizon (120 cm
depth), where the corrosion affecting dolomite and quartz crystals
and the precipitation of calcite are visible. In Fig. 5A the voids, after
quartz dissolution, appear to be filled mainly with calcite. Frequently,
calcification is widespread and calcite nodules occur with dolomite
relicts within the concretions (Fig. 5B). The calcified nodules frequently
had a highly porous texture. The dissolution process of inherited dolo-
mites sometimes began along crystallographic planes and advanced
from inner to outer regions with progressive corrosion, forming hollow
dolomite crystals (Fig. 5C). In addition, corroded dolomite grains
sometimes showed central voids (Fig. 5D). Calcite rhyzocretions also
occurred close to the petrocalcic horizon (Fig. 5D).
Samples from 220 cm in depth showed corroded calcite grains
surrounded by dolomite and clay minerals (Fig. 5E–F). Fig. 5F, corre-
sponding to an inset of Fig. 5E, shows dolomite + clays secondary
products located in a corrosion gulf which includes fragments of
partially altered calcite relics. A detailed observation of precursory
calcite evidenced the presence of a poorly crystalline material, basically
clay minerals, close to irregular corrosion boundaries at the surface of
calcite grains, in addition to inherited chlorites. Clay minerals also
surrounded sub-idiomorphic dolomite crystals with sizes around
2 μm in diameter, and sometimes localized preferentially between
corroded calcite grains and rhombohedral dolomite crystals (Fig. 6A
and B). Poorly defined crystalline material surrounding dolomite
consisted of smectites plus Si-rich amorphous substance (Figs. 6C, 7

Fig. 4. A–D) Histograms of the δ13C ‰ (V-PDB) and δ18O ‰ (V-PDB) of calcites (A and B) and dolomites (C and D). E) Scatter diagram showing the isotopic values (δ13C vs. δ18O) of
the dolomites contained in the fraction b2 mm coming from selected profiles of the studied area. Number data = 72. Note that dolomite samples below 70 cm in depth (full
circles) are highly fractionated. Star corresponds to the average isotopic composition of the inherited dolomites that constitute the fine pebble fraction of the studied soils. The
Meteoric Dolomite Line (MDL) was calculated using the equation for the dolomite-water system (Irwin et al., 1977) with the range of local meteoric water (−8 to −9‰ vs.
V-SMOW) and surface temperatures from 17 to 24 °C.
 J.L. Díaz-Hernández et al. / Chemical Geology 347 (2013) 20–33 27

Fig. 5. A–B) Back-scattered electron images showing precipitation of pedogenic calcite at 120 cm depth (profile 46): Grain of quartz partly corroded and filled with calcite and probably
siderite (A). Calcite nodule including dolomite relicts (B). Note the porous texture of the calcite precipitates in these images. C) Back-scattered electron micrograph of a large idiomorphic
crystal of dolomite in an advanced stage of corrosion located at 120 cm in depth. Note that this corrosion occurs in preferential plains, and progresses from the centre to outer edges.
D) Small dolomite crystals with central voids; in addition, there are pedogenic calcites as rhyzocretions. E) Back-scattered electron image of a partly corroded large calcite grain
surrounded by a mixture of clays and carbonates in a sample from 220 cm in depth (profile 46). F) Inset of E in which we can see some dolomite grains surrounded by clays and corroded
calcite grains.

and Table 2). According to AEM analyses of Table 2 the smectite is inter-
mediate between beidellite and montmorillonite, although several
analyses correspond to mixtures of clay minerals and Si-rich amor-
phous substances. Si excess cannot be incorporated in the smectite
structure and must therefore derive from adjacent non-crystalline
material. Bubbles occurring in relation to clay minerals and amorphous
substance in TEM image (Fig. 6C) resulted from the gradual release of
volatile substances after pervasive electron bombardment. Clay min-
erals were also studied by XRD from separate fractions of the sample
at 220 cm depth (Fig. 8). This study confirms the occurrence of the
following clay minerals: 1) An inherited chlorite is evidenced by (002)
and (004) peaks (Fig. 8A; see also BSE image and qualitative microanal-
ysis of Fig. 6B); 2) a detrital illite with peaks at 1.0 and 0.5 nm d-spacing
corresponding to (001) and (002) reflections; 3) abundant smectites
and minor illite–smectite mixed-layer phases (probably R3-ordered
I/S). The latter is evidenced by the broad basal reflection of the 1.0–
1.2 nm zone in the untreated sample (Fig. 8A). The smectite had a
basal spacing of around 1.4 nm in the air-dried state, which expands
up to 1.8 nm upon ethylene glycol saturation (Fig. 8B).
5. Discussion
5.1. Origin of compositional and mineralogical variations along the
profiles
Organic carbon shows the expected distribution (Fig. 2A and D) for
these types of soils with a usual decrease with depth due to the more in-
tense biogenic activity in the upper horizons. The combination of water
from rainfall, Ca 2+ ions from diverse sources and CO2 generated mostly
from organic-matter decomposition and root respiration (Gile et al.,
1966; Deines, 1980; Rech et al., 2003), the transient decrease in PCO2
and the increase in the concentration of solutes due to evapotranspira-
tion lead to the accumulation in depth of pedogenic calcium carbonate
in arid and semiarid regions. The formation of the petrocalcic horizon
explains the occurrence of the bulge at 40–60 cm depth in the calcite
average regional profile (Fig. 2D), and the calcite enrichment found in
the individual profiles shown in this work (Figs. 2B and 3). Although
the dolomite content in the individual profiles was somewhat irregular
(Fig. 2C), and in particular for depths below 180 cm, it was evident that
28 J.L. Díaz-Hernández et al. / Chemical Geology 347 (2013) 20–33

Fig. 7. A) General TEM image of clays and poorly crystalline material surrounding
dolomite crystals in a sample from 220 cm depth (profile 46). In addition to well-
defined smectite packets (S), with a composition shown in Table 2, amorphous
substance (A) with a mottled appearance and marked dark contrast also occurs in
the poorly crystalline material located around the neoformed dolomite crystal (Dol).
Other areas in the image lacking fringe contrast and of mottled aspect may correspond
to smectites with a slight disorientation of the layers (Sd) or to a mixture of poorly
crystalline smectites and amorphous substance (Table 2). B) Structural features of
the small smectite packets, some of them composed of a few basal layers. The packets
appear to be part of a (gel-like) matrix formed by disoriented smectite-like flakes and a
substance of relatively low atomic number lacking fringe contrast.

a regional decrease occurred near the surface with a marked enrich-

ment in depth (Fig. 2D, Table 1). Moreover, a smooth increase began
at 40 cm and continued to 140 cm depth. Below 180 cm depth, the
values in the profile may have resulted from the interference of

Table 2
Analytical electron microscopy chemical analyses, expressed in atoms per formula unit,
of smectite packets and mixture of smectites and amorphous substance. Normalisation
to 22 charges.

Smectite analyses Mixed analyses

Si 3.79 3.88 3.83 3.78 4.16 4.17 4.12 4.41

AlIV 0.21 0.12 0.17 0.22 −0.16 −0.17 −0.12 −0.41
AlVI 1.55 0.84 1.57 1.69 1.55 1.28 1.43 1.53
Fe(3+) 0.24 0.62 0.17 0.14 0.05 0.28 0.24 0.16
Mg 0.34 0.53 0.22 0.14 0.31 0.48 0.26 0.21
Fig. 6. Electron images of a sample from 220 cm depth (profile 46). A–B) Back-scattered VI
2.14 2.00 1.97 1.97 1.91 2.03 1.93 1.90
electron images of rhombohedral dolomite crystals adjacent to corroded calcite grains and
Ca 0.03 0.03 0.00 0.02 0.05 0.05 0.05 0.03
surrounded by clays. Detrital clays (chlorites) also occur in the area shown in panel B.
K 0.07 0.36 0.34 0.31 0.28 0.16 0.22 0.21
C) Transmission electron microscopy image of a small subhedral dolomite crystal XII
0.10 0.40 0.34 0.33 0.33 0.21 0.28 0.24
surrounded by smectite plus Si-rich amorphous substance.
 J.L. Díaz-Hernández et al. / Chemical Geology 347 (2013) 20–33
Fig. 8. A) Powder X-ray diagram obtained from oriented aggregated (fraction 2) of the sample at 220 cm depth (profile 46), with an indication of d values corresponding to the
reflections of the observed mineral phases. (B) Comparison of air-dried and ethylene-glycol-treated diagrams, where the basal spacing of smectites is displaced from 1.46 nm to
1.76 nm after glycolation.
authigenic processes with the effects of the substratum on the mineral-
ogy. Although the maximum in the case of the dolomite was less marked
than in the calcite, the overall variation in dolomite (around 10%) was
quite similar to the percentage change observed in calcite (around
12%). As occurred for calcite, dolomite dissolution/precipitation may
explain the occurrence of the smooth maximum, just below the
petrocalcic horizon, preceded by the change of sign in the curvature of
the dolomite pattern in the Fig. 2D. Therefore, at the regional scale, do-
lomite underwent superficial dissolution and its precipitation occurred
at greater depths than in the case of calcite. In short, the calcite- and
dolomite-distribution patterns found here clearly indicate that they
were affected by similar processes, but, in the case of dolomite, different
dynamics seem to have occurred within the soil (see below).
Dissolution of the dolomite in the superficial horizon of the studied
profiles (e.g. Fig. 3) in relation to the development of petrocalcic horizon
was also evidenced by Fig. 5A–D, where detrital dolomite appeared
partially or almost completely dissolved and calcite precipitates in
voids and formed rhyzocretions. A general decline in the quartz content
was noted in the profile in Fig. 3, contradicting the assumption that
quartz content remains constant throughout the soil profiles (Lelong,
1969; Souchier, 1971; Sohet et al., 1988). The appearance of a physical
barrier represented by the petrocalcic horizon prevents the redistribu-
tion of the upward-added quartz from diverse sources, mainly aeolian
(Díaz-Hernández et al., 2011), and is therefore responsible for the
observed decrease of quartz below it. However, the quartz content
may decrease below the petrocalcic horizon as a result of dissolution
processes, as evidenced by Fig. 5A, where the precipitation of calcite
after quartz dissolution is evident. Calcite concretions after the dissolu-
tion dolomite and quartz had a marked porous texture (Fig. 5A and B)
29
which may be interpreted in relation to bio-mediated precipitation
processes.
As mentioned above, dolomite enrichment below 40 cm in depth
was observed in the depth-wise variation of dolomite at the regional
scale (Fig. 2D) and in the representative profiles of the Figs. 2C and 3.
One possible explanation would imply no dolomite precipitation, but
only dissolution; thus, the increment in-depth would be due to more
intense dissolution for superficial horizons, the solutes being leached
out of the system. On the other hand, in dissolution/precipitation
mechanism proposed here, dolomite dissolved above 40 cm is poten-
tially precipitated below this depth. Figs. 5E–F and 6, corresponding
to a sample situated at 220 cm in depth in the studied profile, show
the occurrence of dolomite and clay overgrowth in relation to calcite.
Textural relations indicate the dissolution of calcite and the precipita-
tion of dolomite in relation to clays. The specific mechanisms explaining
the observed spatial association of calcite, dolomite and clays in SEM
and TEM images are discussed below.
5.2. Pedogenic isotopic fractionation and neoformed dolomites
The proposed dolomite dissolution/precipitation model is consistent
with the heterogeneity observed in the isotopic data from the dolomite
samples analysed, which showed two different signals (Fig. 4C, D, and
E). Cluster I in Fig. 4E, with higher isotopic values, lies close to the parent
material that supplies the dolomite to the system (star in Fig. 4E). The
precipitation of a neoformed dolomite below the petrocalcic (Figs. 5E,
F and 6) as a result of dissolution of inherited dolomite and pedogenic
calcite of upper horizons explains the high dispersion observed in the
δ13C values of the samples of deeper layers (cluster II in Fig. 4E and
30 J.L. Díaz-Hernández et al. / Chemical Geology 347 (2013) 20–33
smallest left-hand maximum in the histograms of Fig. 4C and D). These
values define the MDL (Meteoric Dolomite Line), a term similar to that
proposed by Lohmann (1988) for the Meteoric Calcite Line (MCL). The
δ13C dispersion is explained by a mixing of inherited carbon from
dolomite-rich parent material, with positive values, and pedogenic
carbon from neoformed dolomite, with very negative δ13C values. In
addition, δ 18O values close to 6‰ (V-PDB) observed for pedogenic dolo-
mites (cluster II in Fig. 4E) are consistent with the δ 18O measured in
surface waters, values ranging from −7.5 and −9‰ (V-SMOW), and
temperatures between ~17 and 24 °C.
The isotopic composition of the dissolved inorganic carbon (DIC)
and oxygen in soil waters below the petrocalcic horizon is in equilib-
rium with the isotopic composition of neoformed dolomites. Carbon
and oxygen are distributed among three chemical species with a
pH-dependent relative abundance (CO2, HCO3− and CO32−). Most of
the carbonic acid derives from the vegetal cover and prompts an
increase of negative δ 13C in DIC (Cerling, 1984, 1991; Usdowski and
Hoefs, 1993; Reyes et al., 1998). In the pedogenic medium studied,
this isotopic signature is transported towards deeper layers through
carbonate dissolution (mostly calcite) above the petrocalcic horizon.
The negative carbon isotope composition of the neoformed dolomite
is determined mainly by the isotopic pedogenic calcite because pore
fluids generally have low carbon contents so that the 13C of the pre-
cursor is generally retained (Hoefs, 2004). Low δ 13C and δ 18O values
are expected in biomediated calcite and dolomite precipitates (Jones
and Donnelly, 2004; Wacey et al., 2007). Multiple cyclic dissolution/
precipitation processes that affected the carbonates in the soils
explain not only the formation of an isotopically light calcite, related
to pedogenic–biogenic processes above the petrocalcic horizon
(Magaritz and Amiel, 1980; Magaritz et al., 1981), but also the forma-
tion of an isotopically light dolomite, occurring below the petrocalcic
horizon, in which the isotopic composition differs from that observed
for the inherited dolomites (Fig. 4). The isotopic values of the pedogenic
calcite co-existing with neoformed dolomite may be assigned to the
maximum observed in the histograms of Fig. 4A–B. Calcite isotopic
values are more negative than those of dolomite, as expected, because
most of the calcite in soil is pedogenic in origin. The pedogenic dolomite
precipitated in depth has a marked isotopic contribution from inherited
dolomite, and therefore lower negative values than those measured for
the pedogenic calcite are expected.
Dolomite isotopic composition is not usually determined in soils,
and much less in deep layers (“subsoil”). Until now, the occurrence
of a pedogenic 13C (and 18O) signal in soils is associated exclusively
with the calcite genesis. In the case studied, the neoformed dolomite,
characterized by relatively negative values, forms after partial dissolu-
tion of the calcite, with a partial “transference” of the biogenic isotopic
signature. Therefore, calcite precipitation is interpreted here as an inter-
mediate stage for dolomite precipitation, below the petrocalcic horizon,
because of the carbonate dissolution–precipitation cycles forming pedo-
genic calcite are also responsible of the formation of the pedogenic
dolomites.
Kinetic effects, such as diffusion transport, may also induce a
certain fractionation in soils. The carbon isotope fractionation in soils
caused by diffusion processes has been estimated to be around 4‰
(Cerling, 1984; Hesterberg and Siegenthaler, 1991). According to
these authors 13CO2 diffuses into the atmosphere at lower rates than
the lighter 12CO2 molecule. However this effect is clearly counteracted
by the major 12CO2 production related to root respiration and organic-
matter decomposition in the upper horizons of soil, as evidenced clearly
by the lower δ13C values recorded for the pedogenic calcite of soil in
relation to the pedogenic dolomite occurring in the subsoil. The kinetic
effect discussed above resulting in 13C enrichment would partly explain
the shift from the very negative δ 13C at the C3 vegetal cover (~−27‰
vs. V-PDB, Deines, 1980) due to CO2 from organic matter decomposi-
tion, towards values around −12‰ in the soil (Cerling, 1984, 1991;
Reyes et al., 1998). Other kinetic processes such as intracrystalline
diffusion of C and O within the dolomite structure might explain isotope
exchange between the solid and the pore fluid in soils, and the observed
fractionation in Fig. 4, without the need for an explanation involving
the dolomite dissolution–precipitation process. However, alternative
explanations of these isotopic data based on intracrystalline diffusion
processes should be discarded in our case because of: 1) According to
the Arrhenius relationship between the diffusion rate and the tempera-
ture, intracrystalline diffusion is negligible at low temperature (Putnis,
2003). 2) Volume diffusion of C within the structure of the rhombohe-
dral carbonates occurs by diffusion of carbonate anions, whereas the
lattice diffusion of O is due to vacancy migration (Labotka et al.,
2000); this explains why the diffusivity of carbon is 4 orders of magni-
tude smaller than the diffusivity of oxygen in calcite (Ferry et al., 2010).
The latter point disagrees with our observations, where a broader range
of variation is found for δ 13C compared to that of δ 18O for the cluster II
(Fig. 4E) corresponding to the dolomite located at deeper levels.
5.3. Physico-chemical constraints on the dolomite formation in the
subsoil
As shown, dissolution of inherited dolomite above petrocalcic con-
stitutes the main source for the Mg 2+ enrichment in the soil solution
necessary for dolomite neoformation. An exception to this is provided
by the profiles with altered ophitic rocks in their subsoil, where minor
dolomite was formed at great depth (below 200 cm, Fig. 2C). The
montmorillonitic component of the smectites localized preferentially
between corroded calcite grains and rhombohedral dolomite crystals
(Table 2 and Fig. 6A–B) may constitute an additional source for Mg 2+.
Despite the occurrence of high Mg:Ca mole ratio, Mg rich calcite form
only at very high salinity (Stanley and Hardie, 1998). However, for
medium salinities, a low Mg content was expected in the calcite formed
from solutions in soils. This is because Mg2+ ions tend to remain in
solution during carbonate precipitation, where they are thought to be
heavily solvated (de Leeuw and Parker, 2001), due to strongly attractive
Mg–water interactions.
The formation of the petrocalcic horizon prevents rain-water infil-
tration into the subsoil. The observed dolomite enrichment in the
subsoil probably occurred during the initial stages of the formation
of the petrocalcic horizon (soft caliches), and before its hardening
that led to the consolidation of the impermeable barrier. The high
content of gravels and sands throughout the profile (Fig. 2E) favoured
rainwater infiltration into the subsoil. Under these physical conditions,
evapotranspiration processes affected not only the soil, but also deeper
levels during the less rainy climatic episodes. The decreased water
content in the subsoil diminished bulk flow of the grain-boundary
fluid containing dissolved solutes. Under this regime, the fluid film is
restricted to a monolayer of molecules adsorbed onto the smectite
particles. To understand the role of the water molecules adsorbed
onto smectites in the incorporation of Mg to the structure of the calcite,
we dealt here with the transport of solute in a viscous solid media
(Putnis et al., 1995; Fernández-Díaz et al., 1996; Sánchez-Navas et al.,
2009). It can be deduced from Table 3 that the quantity of Mg incorpo-
rated into the structure of the calcite increased in viscous medium and
that the incorporation of Mg into magnesian-calcite produced a lattice
contraction of the rhombohedral carbonate, as well as a reduction of
Table 3
X-ray diffraction results for magnesian calcites obtained in batch precipitation experi-
ments by using two types of media. In both cases the concentrations of the mixed soluble
salts are the same: 50 mM Na2CO3, 50 mM CaCl2·6H2O, and 0.1 M MgCl2·6H2O (starting
pH 7.4).
Type of medium d104 (Å) FWHM (° 2θ)a % Mg
Liquid 3.01 0.3 10
Viscous 2.95 1.26 25
a
FWHM: full width at half maximum (peak broadening measured for 104 reflection
in the XRD diagrams of Fig. 10).
 J.L. Díaz-Hernández et al. / Chemical Geology 347 (2013) 20–33 31

crystallinity. In addition, metastable phases also appeared when a

viscous medium was used for the precipitation experiment (Fig. 9).
Micro-textural data (Fig. 6) indicate that the precipitation of neoformed
dolomite in depth after partial dissolution of pedogenic calcite was
mediated by the smectites of the soils studied (Figs. 7 and 8). Clay min-
erals surrounding neoformed dolomite crystals acted as a catalytic
material in the process of incorporation of Mg into the structure of the
rhombohedral carbonates. This was because the transport of reacting
material from dissolution sites to those of growth was the rate-
determining process. Therefore, as suggested by our laboratory experi-
ments, it is necessary to consider the constraints that transport proper-
ties may make on the incorporation of an element during crystal
growth. Fig. 10A shows the concentration profile of Mg at the crystal–
water interface during interface-controlled calcite growth. As suggested
at the beginning of this section, the partition coefficient between the
concentration of magnesium in carbonate crystal and the concentration
of magnesium in bulk medium, [Mg]carb/[Mg]m, must be clearly less
than 1. For a diffusion-controlled growth of dolomite, the concentration
of Mg rises above [Mg]m at the crystal–medium interface ([Mg]im). If Fig. 10. Concentration profiles of Mg at the crystal–medium interface. A) Concentration
the partition coefficient remains constant, [Mg]carb increases with distribution at an infinitely fast bulk-diffusion rate. B) Steady-state distribution at very
decreasing diffusion rate until, in the limit, [Mg]carb = [Mg]m. As a slow surface-diffusion rate along the grain boundaries of clay minerals. ([Mg]carb,
concentration of magnesium in carbonate crystal; [Mg]m, concentration of magnesium
result the effective partition coefficient equals 1, and therefore most of
in bulk medium; [Mg]im, concentration of magnesium at the carbonate crystal–medium
the Mg of the medium is then incorporated into the structure of the interface; Δ, thickness of the diffusion boundary layer). Based on Henderson (1986).
rhombohedral carbonate (Fig. 10B).
Previous experimental data and the occurrence in soils of high-
magnesium calcite together with well-ordered dolomite (Whipkey Before the formation of nonpermeable petrocalcic horizon, the
and Hayob, 2008) suggest that our neoformed dolomite that precipitated high contents in gravel and sand fractions in the studied soils
in relation to smectites in the “subsoil” should be, at least initially, a favoured infiltration and evapotranspiration processes in the subsoil.
“protodolomite” or “Mg-kutnohorite” with essentially disordered Ca– The moisture in the subsoil was a determining factor controlling the
Mg distribution. precipitation of Ca and Mg carbonates, the precursors of dolomite.
The precipitation of the neoformed dolomite below the petrocalcic
6. Conclusions horizon was related to the presence of smectites. Smectite retained
soil water and acted as a catalyst for the incorporation of Mg into
The present study detected the presence of a clear enrichment of the carbonate.
dolomite beyond the lower limit of biological activity in semiarid The geochemical model proposed here for the incorporation of Mg
soils. This limit coincides with the rooting depth of native perennial to the carbonate in slow diffusive transport medium, such as a clayey-
plants and, in the case study, is defined by the occurrence of the rich and water deficient environments, would also explain the close
petrocalcic horizon. association of Mg-carbonates with clays observed on the surface of
Isotopic data establish a bimodal distribution of δ 13C and δ 18O for Mars (Ehlmann et al., 2008).
dolomites, indicating the occurrence of an inherited and a neoformed This study documents the occurrence of changes in carbonate
dolomite. Isotopically light dolomite formed below the petrocalcic mineralogy in depth. Levels below the petrocalcic are traditionally
horizon after partial dissolution of the pedogenic calcite, which trans- considered as zones not affected by the pedogenic processes.
ferred the pedogenic (biological) 13C and 18O signature from the soil In the soils studied, the dolomite content reached 43 kg m−2
(s. str.) to subsoil. (Díaz-Hernández, 2010), and therefore dolomite should be considered

Fig. 9. X-ray diffraction diagrams for magnesian calcites obtained in batch precipitation experiments (see Section 3.7 and Table 3) by using viscous (A) and liquid media (B). d104
spacing of calcite varies from 0.301 nm for the liquid medium to 0.295 nm for the viscous media. In the viscous medium monohydrocalcite also precipitates together with the
magnesian calcite.
32 J.L. Díaz-Hernández et al. / Chemical Geology 347 (2013) 20–33
in the quantification of carbon stored in arid–semiarid soils. Pedogenic
processes occurring in the subsoil, such as the dolomite formation,
should be included in the carbon-soil balances.
Acknowledgements
We acknowledge support from grant P11-RNM-7067 of the Junta de
Andalucía (C.E.I.C.—S.G.U.I.T.) and the projects CGL2007-65572-C02-01/
BTE, CGL-2009-09249, CGL2010-21257-C02-01 and CGL2012-32169 of
the Ministerio de Ciencia e Innovación (MICINN, Spain). We would like
to thank Drs. D. Martín-Ramos and F. Nieto for their help with the XRD
analyses, A. González, M.M. Abad and M.J. Guerrero for their guidance
with SEM and TEM studies and Dr. A. Delgado for his assistance with
the stable isotope data. We thank the anonymous reviewers for their sug-
gestions, which contributed to improvement the manuscript. Mr. David
Nesbitt revised the English manuscript.
References
Al-Aasm, I.S., Taylor, B.E., South, B., 1990. Stable isotope analysis of multiple carbonate
samples using selective acid extraction. Chemical Geology 80, 119–125.
Arkley, R.J., 1963. Calculation of carbonate and water movement in soil from climatic
data. Soil Science 96, 239–248.
Arvidson, R.S., Mackenzie, F.T., 1999. The dolomite problem: control of precipitation
kinetics by temperature and saturation state. American Journal of Science 299,
257–288.
Botha, G.A., Hughes, J.C., 1992. Pedogenic palygorskite and dolomite in a Late Neogene
sedimentary succession, nothwestern Transvaal, South Africa. Geoderma 53,
139–154.
Bui, E.N., Loeppert, R.H., Wilding, L.P., 1990. Carbonate phases in calcareous soils of the
western United States. Soil Science Society of America Journal 54, 520–529.
Capo, R.C., Whipkey, C.E., Blachère, J.R., Chadwick, O.A., 2000. Pedogenic origin of
dolomite in a basaltic weathering profile, Kohala peninsula, Hawaii. Geology 28,
271–274.
Cerling, T.E., 1984. The estable isotopic composition of modern soil carbonates and its
relationships to climate. Earth and Planetary Science Letters 71, 229–240.
Cerling, T.E., 1991. Carbon dioxide in the atmosphere: evidence from cenozoic and
mesozoic paleosols. American Journal of Science 291, 377–400.
Cerling, T.E., Quade, J., 1993. Stable carbon and oxygen isotopes in soil carbonates. In:
Swart, P.K., Lohmann, K.C., Mckenzie, J., Savin, S. (Eds.), Climate Change in Continental
Isotopic Records. Geophysical Monograph, vol. 78. American Geophysical Union,
Washington, pp. 217–231.
de Leeuw, N.H., Parker, S.C., 2001. Surface–water interactions in the dolomite problem.
Physical Chemistry Chemical Physics 3, 3217–3221.
Deelman, J.C., 2005. Low Temperature Formation of Dolomite and Magnesite. Geology
Series, Version 2.1. Compact Disc Publication, Eindhoven.
Deines, P., 1980. The isotopic composition of reduced carbon. In: Fritz, A., Fontes, P.
(Eds.), The Terrestrial Environment. Handbook of Environmental Isotope
Geochemistry. Elsevier Scientific Press, pp. 329–434.
Deleuze, M., Brantley, S.L., 1997. Inhibition of calcite crystal growth by Mg2+ at 100 °C
and 100 bars: influence of growth regime. Geochimica et Cosmochimica Acta 61,
1475–1485.
Díaz-Hernández, J.L., 2010. Is carbon storage in soils underestimated? Chemosphere
80, 346–349.
Díaz-Hernández, J.L., Juliá, R., 2006. Geochronological position of badlands and
geomorphological patterns in the Guadix–Baza basin (SE Spain). Quaternary
Research 65, 467–477.
Díaz-Hernández, J.L., Barahona-Fernández, E., Linares-González, J., 2003. Organic and
inorganic carbon in soils of semiarid regions: a case study from the Guadix–Baza
basin (Southeast Spain). Geoderma 114, 65–80.
Díaz-Hernández, J.L., Martín-Ramos, J.D., López-Galindo, A., 2011. Quantitative analysis
of mineral phases in atmospheric dust deposited in the south-eastern Iberian
Peninsula. Atmospheric Environment 45, 3015–3024.
Drees, L.R., Wilding, L.P., 1987. Micromorphic record and interpretations of carbonate
forms in the rolling plains of Texas. Geoderma 40, 157–175.
Ehlmann, B.L., Mustard, J.F., Murchie, S.L., Poulet, F., Bishop, J.L., Brown, A.J., Calvin,
W.M., Clark, R.N., Marais, D.J., Milliken, R.E., Roach, L.H., Roush, T.L., Swayze, G.A.,
Wray, J.J., 2008. Orbital identification of carbonate-bearing rocks on mars. Science
322, 1828–1832.
FAO-UNESCO, 1988. Soil Map of the World. Revised Legend. FAO, Roma.
Fernández-Díaz, L., Putnis, A., Prieto, M., Putnis, C.V., 1996. The role of magnesium in
the crystallization of calcite and aragonite in a porous medium. Journal of
Sedimentary Research 66, 482–491.
Ferry, J.M., Ushikubo, T., Kita, N.T., Valley, J.W., 2010. Assessment of grain-scale homo-
geneity and equilibration of carbon and oxygen isotope compositions of minerals
in carbonate-bearing metamorphic rocks by ion microprobe. Geochimica et
Cosmochimica Acta 74, 6517–6540.
Fritz, P., Smith, D.G.W., 1970. The isotopic composition of secondary dolomites.
Geochimica et Cosmochimica Acta 34, 1161–1173.
García-Aguilar, J.M., Martín, J.M., 2000. Late Neogene to recent continental history and
evolution of the Guadix–Baza basin (SE Spain). Revista de la Sociedad Geológica de
España 13, 65–77.
Ghebre-Egziabhier, K., St. Arnaud, R.J., 1983. Carbonate mineralogy of lake sediments
and surrounding soils. 2. The Qu'Appelle Lakes. Canadian Journal of Soil Science
63, 259–268.
Gibert, L., Ortí, F., Rosell, L., 2007. Plio-Pleistocene lacustrine evaporates of the Baza
Basin (Betic Chain, SE Spain). Sedimentary Geology 200, 89–116.
Gile, L.H., Peterson, F.F., Grossman, R.B., 1966. Morphological and genetic sequences of
carbonate accumulation in desert soils. Soil Science 1015, 347–360.
Goddard, M.A., Mikhailova, E.A., Post, C.J., Schlautman, M.A., 2007. Atmospheric Mg2+
wet deposition within the continental United States and implications for soil
inorganic carbon sequestration. Tellus Series B 59, 50–56.
Goldsmith, J.R., Graf, D.L., 1958. Structural and compositional variations in some
natural dolomites. Journal of Geology 66, 678–693.
Hardie, L.A., 1987. Perspectives dolomitization: a critical view of some current views.
Journal of Sedimentary Petrology 57, 166–183.
Henderson, P., 1986. Inorganic Geochemistry. Pergamon Press, Oxford (356 pp.).
Hesterberg, R., Siegenthaler, U., 1991. Production and stable isotopic composition of
CO2 in a soil near Bern Switzerland. Tellus 43B, 197–205.
Hoefs, J., 2004. Stable Isotope Geochemistry. Springer-Verlag, Berlin Heidelberg (244
pp.).
IGME (1973–1999) Mapa Geológico de España, escala 1:50.000, Hojas 929, 930, 950,
951, 972, 973, 993, 994, 1011 and 1012.
Irwin, H., Curtis, C., Coleman, M., 1977. Isotopic evidence for source of diagenetic
carbonates formed during burial of organic-rich sediments. Nature 269, 209–213.
Jobággy, E.G., Jackson, R.B., 2000. The vertical distribution of soil organic carbon and its
relation to climate and vegetation. Ecological Applications 10, 423–436.
Jones, M.B., Donnelly, A., 2004. Carbon sequestration in temperate grassland ecosystems
and the influence of management, climate and elevated CO2. New Phytologist 164,
423–439.
Kim, S.T., O'Neil, J.R., 1997. Equilibrium and nonequilibrium oxygen isotope effects in
synthetic carbonates. Geochimica et Cosmochimica Acta 61, 3461–3475.
Kohut, C., Dudas, M.J., Muehlenbachs, K., 1995. Authigenic dolomite in a saline soil in
Alberta, Canada. Soil Science Society of America Journal 59, 1499–1504.
Labotka, T.C., Cole, D.R., Riciputi, L.R., 2000. Diffusion of C and O in calcite at 100 MPa.
American Mineralogist 85, 488–494.
Lelong, F., 1969. Nature et gènèse des produits d'altération de roches cristallines sous
climat tropical humide (Guyane Française). Mèmoires Science de la Terre, Nancy,
14, 187 pp.
Loeppert, R.H., Suárez, D.L., 1996. Carbonate and gypsum. In: Sparks, D.L. (Ed.),
Methods of Soil Analysis, Part 3 Chemical Analysis. SSSA Book series, No 5. ASA &
SSSA, Madison, WI, pp. 437–474.
Lohmann, K.Z., 1988. Geochemical Patterns of Meteoric Diagenetic Systems and their
Application to Studies of Paleokarst. In: James, N.P., Choquette, P.W. (Eds.),
Paleokarst. Springer-Verlag, Berlin, pp. 58–80.
Lorimer, G.W., Cliff, G., 1976. Analytical electron microscopy of minerals. In: Wenk, H.R.
(Ed.), Electron Microscopy in Mineralogy. Springer-Verlag, Berlin, pp. 506–519.
Magaritz, M., Amiel, A.J., 1980. Calcium carbonate in a calcareous soil from the Jordan
Valley, Israel: its origin as revealed by the stable carbon isotope method.
Soil Science Society of America Journal 44, 1059–1062.
Magaritz, M., Kaufman, A., Yaalon, D.H., 1981. Calcium carbonate nodules in soils: 18O/16O
and 13C/12C ratios and 14C contents. Geoderma 25, 157–172.
MAPA, 1989. Caracterización agroclimática de la Provincia de Granada. Ed. Secretaría
General Técnica del MAPA. 197 pp.
Martín-Ramos, J.D., 2004. Using XPowder®, A Software Package for Powder X-ray
Diffraction Analysis. D.L.GR-1001/0484-609-1497-6 (Spain).
McCrea, J.M., 1950. On the isotopic chemistry of carbonates and a paleotemperature
scale. Journal of Chemical Physics 18, 849–857.
Mikhailova, E.A., Post, C.J., 2006. Effects of land use on soil inorganic carbon stocks in
the Russian Chernozem. Journal of Environmental Quality 35, 1384–1388.
Mikhailova, E.A., Post, C.J., Cihacek, L., Ulmer, M., 2009. Soil inorganic carbon sequestration
as a result of cultivation in the mollisols. In: McPherson, B.J., Sundquist, E.T. (Eds.),
Carbon Sequestration and Its Role in the Global Carbon Cycle, Geophysical Monograph
Series 183. A.G.U., Washington, D.C., pp. 129–133.
Ming, D.W., 2002. In: Lal, R. (Ed.), Encyclopedia of Soil Science. Marcel Dekker,
New York, pp. 139–141.
Mucci, A., Morse, J.W., 1983. The incorporation of Mg2+ and Sr2+ into calcite
overgrowths: influences of growth rate and solution composition. Geochimica
et Cosmochimica Acta 47, 217–233.
Peña, J.A., 1985. La Depresión de Guadix–Baza. Estudios Geológicos 41, 33–46.
Podwojewski, P., 1995. The occurrence and interpretation of carbonate and sulphate
minerals in a sequence of Vertisols in New Caledonia. Geoderma 65, 223–248.
Putnis, A., 2003. Introduction to Mineral Sciences. Cambridge University Press (457 pp.).
Putnis, A., Prieto, M., Fernández-Díaz, L., 1995. Fluid supersaturation and crystallization
in porous media. Geological Magazine 132, 1–13.
Rech, J.A., Quade, J., Hart, W.S., 2003. Isotopic evidence for the source of Ca and S in soil
gypsum, anhydrite and calcite in the Atacama Desert, Chile. Geochimica et
Cosmochimica Acta 67, 575–586.
Reeder, R.J., Sheppard, C.E., 1984. Variation in lattice parameters in some sedimentary
dolomites. American Mineralogist 69, 520–527.
Reyes, E., Pérez del Villar, L., Delgado, A., Cortecci, G., Núñez, R., Pelayo, M., Cózar, J.,
1998. Carbonatation processes at the El Berrocal analogue granitic system
(Spain): mineralogical and isotopic study. Chemical Geology 150, 293–315.
Rosenbaum, J., Sheppard, S.M.F., 1986. An isotopic study of siderites, dolomites and an-
kerites at high temperatures. Geochimica et Cosmochimica Acta 50, 1147–1150.
 J.L. Díaz-Hernández et al. / Chemical Geology 347 (2013) 20–33
Salomons, W., Mook, W.G., 1976. Isotope geochemistry of carbonate dissolution and
reprecipitation in soils. Soil Science 122, 15–24.
Salomons, W., Goudie, A., Mook, W.G., 1977. Isotopic composition of calcrete deposits
from Europe, Africa and India. Earth Surface Processes and Landforms 3, 43–57.
Sánchez-Navas, A., Martín-Algarra, A., Rivadeneyra, M.A., Melchor, S., Martín-Ramos,
J.D., 2009. Crystal-growth behaviour in Ca–Mg carbonate bacterial spherulites.
Crystal Growth & Design 9, 2690–2699.
Shermann, D.G., Shultz, F., Always, F.J., 1962. Dolomitization of soils of the Red River
Valley, Minnesota. Soil Science 94, 304–313.
Sobecki, T.M., Karathanasis, A.D., 1987. Quantification and compositional characterization
of pedogenic calcite and dolomite in calcic horizons of selected Aquolls. Soil Science
Society of America Journal 51, 683–690.
Sohet, K., Herbauts, J., Gruber, W., 1988. Changes caused by Norway spruce in ochreous
brown earth, assessed by isoquartz method. Journal of Soil Science 39, 549–561.
Soil Survey Staff, 1999. Soil taxonomy, a basic system of soil classification for making
and interpreting soil surveys. U.S. Department of Agriculture, Soil Conservation
Service, Agriculture Handbook No. 436, Washington, D.C. (754 pp.).
Souchier, B., 1971. Evolution des sols sur roches cristallines à l'étage montagnard
(Vosges). Mémoires du Service de la Carte Géologique d'Alsace et de Lorraine 33,
134 pp.
St. Arnaud, R.J., 1979. Nature and distribution of secondary soil carbonates within
landscapes in relation to soluble Mg2+/Ca2+ ratios. Canadian Journal of Soil Science
59, 87–98.
St. Arnaud, R.J., Herbillon, A.J., 1973. Occurrence and genesis of magnesium-bearing
calcites in soils. Geoderma 9, 279–298.
Stanley, S.M., Hardie, L.A., 1998. Secular oscillations in the carbonate mineralogy of
reef-building and sediment-producing organisms driven by tectonically forced
33
shifts in seawater chemistry. Palaeogeography Palaeoclimatology Palaeoecology
144, 3–19.
Stevenson, B.A., Kelly, E.F., McDonald, E.V., Busacca, A.J., 2005. The stable carbon
isotope composition of soil organic carbon and pedogenic carbonates along a
bioclimatic gradient in the Palouse region, Washington State, USA. Geoderma
124, 37–47.
USDA-NRCS, 1996. Soil survey laboratory methods manual. S.S. Investigation Report
No. 42, Version 3.0 (716 pp.).
Usdowski, E., Hoefs, J., 1993. Oxygen isotope exchange between carbonic acid, bicarbonate,
carbonate, and water: a re-examination of the data of McCrea (1950) and an expres-
sion for the overall partitioning of oxygen isotopes between the carbonate species
and water. Geochimica et Cosmochimica Acta 57, 3815–3818.
Vera, J.A., 1970. Estudio estratigráfico de la Depresión de Guadix–Baza. Boletín
Geológico y Minero 81, 429–462.
Wacey, D., Wright, D.T., Boyce, A.J., 2007. A stable isotope study of microbial dolomite
formation in the Coorong region, South Australia. Chemical Geology 244, 155–174.
Whipkey, Ch.E., Hayob, J.L., 2008. Textural and compositional evidence for the evolution of
pedogenic calcite and dolomite in a weathering profile on the Kohala Peninsula,
Hawaii. Carbonate Evaporite 23, 104–112.
Wright, V.P., 1992. Paleosol recognition: a guide to early diagenesis in terrestrial
settings. In: Wolf, K.H., Chilingarian, G.V. (Eds.), Diagenesis, III: Amsterdam
(Elsevier): Developments in Sedimentol, 47, pp. 591–619.
Wright, D.T., 1999. The role of sulphate-reducing bacteria and cyanobacteria in dolomite
formation in distal ephemeral lakes of the Coorong region, South Australia. Sedimen-
tary Geology 126, 147–157.
Wright, V.P., Tucker, M.E., 1991. Calcretes. Ed. Blackwell, London (352 pp.).