ELSEVIER Earth and Planetary Science Letters 163 (1998) 69–81
Chronology of the late Turolian deposits of the Fortuna basin
(SE Spain): implications for the Messinian evolution of
the eastern Betics
Miguel Garce´s a,b, , Wout Krijgsman b , Jorge Agusti´ c
a Grup de Geodina` mica i Ana`lisi de Conques, Dept. d’Estratigrafia i Paleontologia, Universitat de Barcelona,
Marti ´ i Franque`s s=n, 08028 Barcelona, Spain
b Paleomagnetic Laboratory Fort Hoofddijk, Budapestlaan 17, 3584 CD Utrecht, Netherlands
c Institut de Paleontologia M. Crusafont, Escola Industrial 23, 08201 Sabadell, Spain
Received 30 March 1998; revised version received 7 July 1998; accepted 11 August 1998
Abstract
The magnetostratigraphy of the mammal-bearing alluvial fan–fan delta sequences of the Fortuna basin (SE Spain) has
yielded an accurate chronology for the late Turolian (Messinian) basin infill. From early to late Messinian (at least between
6.8 and 5.7 Ma), the Fortuna basin records the sedimentation of alluvial–palustrine deposits over a confined shallow
basin. Changing environmental conditions in the latest Messinian are illustrated by the retreat of palustrine facies. A rapid
progradation of the marginal clastic wedges and the initiation of an efficient basin drainage at . 5.8 Ma (lower part of
chron C3r) most likely represents the onshore response to the drastic drop of base level taking place during the Messinian
salinity crisis. This study further provides improved age estimates for the late Turolian land mammal events in southern
Spain. The oldest MN 13 locality in the studied sections is correlated to chron C3Ar at an age of 6.8 Ma. The entry
of camels and the murid Paraethomys in southern Spain occurs in chron C3An.1n at 6.1 Ma, and gives further support
for land mammal exchange between Africa and the Iberian peninsula prior to the salinity crisis, in good agreement with
results from northern Africa [M. Benammi, M. Calvo, M. Pre´vot, J.J. Jaeger, Magnetostratigraphy and paleontology of Ai¨t
Kandoula basin (High Atlas, Morocco) and the African–European late Miocene terrestrial fauna exchanges, Earth Planet.
Sci. Lett. 145 (1996) 15–29]. The age of the studied sequences provides important constraints on the understanding of
the sedimentary evolution of the eastern Betic margin, and shows that previous interpretations of the evaporitic–diatomitic
sequences of the Fortuna basin, as being coeval to the late Messinian salinity crisis in the Mediterranean, are not correct.
The confinement leading to the emergence of the Fortuna basin occurred in the late Tortonian to earliest Messinian, similar
to other intramontane basins in the Betics. Therefore, the inclusion of the Fortuna basin in a hypothetical marine Betic
Corridor during the late Messinian is no longer tenable. Ó 1998 Elsevier Science B.V. All rights reserved.
Keywords: Spain; Betic Zone; Messinian; Turolian; magnetostratigraphy; biostratigraphy; mammals
 Corresponding author. Tel.: C 34 93 402 1395=76; Fax: C 34 93 402 1340; E-mail: garces@natura.geo.ub.es
0012-821X/98/$ – see front matter Ó 1998 Elsevier Science B.V. All rights reserved.
PII: S0012-821X(9 8 )0 0 1 7 6 -9.70 M. Garce´s et al. / Earth and Planetary Science Letters 163 (1998) 69–81
1. Introduction
The Neogene basins in the eastern Betic Ranges
(SE Spain) offer a privileged setting to study the
evolution of the western Mediterranean basin during
the late Miocene and Pliocene [1–3]. Two marine
gateways connected the western Mediterranean to
the Atlantic in late Tortonian times, the Rifian Cor-ridor
through Morocco and the North Betic Corridor
through Spain. Ongoing tectonic processes, and a su-perimposed
paleoclimatic (glacio-eustatic sea-level
fluctuations) component, progressively closed these
gateways during the Messinian and caused a com-plete
isolation of the Mediterranean. The result was
a series of successive events that changed the pale-ogeography
and paleoceanography of the Mediter-ranean.
The best known, but least understood, event
is the ‘Messinian salinity crisis’, a period during
which very thick evaporite units were deposited all
over the Mediterranean, followed by continental and
brackish water sediments which are related to the
desiccation phase of the latest Messinian. Paleogeo-graphic
reconstructions of these marine gateways,
as well as the precise age of closure, are still un-certain
because tectonic deformation and large-scale
olistolith bodies, covering the youngest marine sedi-ments,
hamper a reliable age control.
Paleogeographic reconstructions of the eastern
Betics [3] indicates that the Fortuna basin (Fig. 1)
occupied the northeastern part of the Betic Cor-ridor.
The sedimentary infill of the Fortuna basin
[1,4–6] can be grouped into three main units: (1) a
thick marine sequence consisting of turbidites and
pelagic marls (Los Ban˜os Formation [4]), informally
referred to as Fortuna marls [1], (2) a regressive
evaporitic sequence [7] with gypsiferous marls, di-atomites,
massive gypsum and conglomerates (Rio
Chicamo Formation [4]), and (3) a post-evaporitic
basin infill consisting of a mainly continental alluvial
to lacustrine succession (Rambla Salada Formation
[4]). The regressive sequence from marine marls to
evaporites and continental deposits has been inter-preted
in the context of sea level lowering that led
to the Messinian salinity crisis [6]. Furthermore, de-tailed
correlations were proposed between the Rio
Chicamo evaporites and the well-known gypsum de-posits
(Lower Evaporites) of the Mediterranean basin
[4,8]. Based on these correlations, Mu¨ ller and Hsu¨
[4] claimed that the marine connection through the
Betic corridor was open till the latest Messinian.
These correlations and interpretations were not
substantiated, however, by a better resolved chronol-ogy.
Concerning the age of the formations there
have been conflicting interpretations. While early
datings suggested a latest Tortonian age for the evap-orites
of the Fortuna basin [1], a revised plank-tonic
biostratigraphy led Lukowski and colleagues
[9] to date the Fortuna marls as Messinian. Conse-quently,
the overlying evaporites were considered as
late Messinian. However, the postulated Messinian
age for the Fortuna marls is not validated by any
convincing biostratigraphic marker. The FAD of
Globorotalia conomiozea denotes the lower bound-ary
of the Messinian Stage, but this species is not
found in the pre-evaporitic planktonic-rich facies of
the central part of the basin [9,10], while it is cited
in the shallow deposits overlying the onset of the
evaporites and confined conditions [1,5].
In contrast to the marine units, the continental
sequences of the Fortuna basin provide a reliable and
detailed biostratigraphic record. Several rich, late
Turolian, vertebrate assemblages [11–14] indicate a
middle to late Messinian age for these sediments.
This age is supported by marine–continental corre-lations
of classical localities in neighbouring regions
[15,16]. In this paper we present the magnetostratig-raphy
of the late Turolian alluvial fan–fan delta
sequences which crop out along the SE margin of
the Fortuna basin. The results of this study pro-vide
an accurate chronology for the late Turolian
(Messinian) bioevents in the western Mediterranean,
as well as important constraints on the sedimentary
evolution of the east Betic margin.
2. Geological setting of the Fortuna basin
The pre-Neogene rocks of the eastern Betics are
divided into two units (Fig. 1). The External Bet-ics
consist of the deformed southeastern margin of
the Iberian Massif and the Internal Betics consist of
the allochtonous units which were incorporated into
the orogen after a westward drift during the middle
Miocene. A number of intramontane basins devel-oped
in the eastern Betics since the early Tortonian
[3,17,18] under an overall NNW–SSE compressional.M. Garce´s et al. / Earth and Planetary Science Letters 163 (1998) 69–81 71
Fig. 1. (A) Geological sketch of the Neogene basins in the Betic Ranges. The Granada, Guadix–Baza and Fortuna basins, lying on
the contact between the External and Internal Zones, formed part of a North Betic corridor [3] which connected the Atlantic and the
Mediterranean in late Tortonian times. (B) The Fortuna basin in the eastern Betics is bounded by two major shear zones, the Crevillente
and the Alhama de Murcia faults. The Carrascoy and Orihuela–Callosa massifs are relicts of its south-eastern margin. SL: Sifo´n de
Librilla section; CH: Chorrico section; AMF: Alhama de Murcia Fault; M: Murcia; F: Fortuna; L: Librilla.
regime [3,19] as a result of the Africa–Europe plate
collision.
The Fortuna basin (Fig. 1) lies on the contact
between the External and Internal Betics and is
bounded by two major NE–SW shear zones. The
North Betic–Crevillente fault zone [20] delineates
the tectonically active northwestern basin margin,
which acted as the main sediment source. The Al-hama
de Murcia wrench fault [21,22] lies along the
Guadalentin corridor and approximately outlines the
SE margin of the Fortuna basin. Late Neogene N–S
compression resulted in important left-lateral shear
and basement uplift along the basin margins. An
emerged massif to the East of the Alhama de Murcia
Fault [1,21] acted as an active source area for a set of
alluvial fans which spread into the basin during the
late Miocene. Relics of this paleorelief are the Car-rascoy,
Orihuela and Callosa massifs (Fig. 1), which
mainly consist of Permotriassic metasediments and
volcanic rocks of the Internal Betics. The tectonic
uplift of the southeastern margin played an impor-tant
role in the initiation of restricted environments
and the subsequent isolation of the Fortuna basin
from the Mediterranean basin [5].
3. Studied sections
The present study focuses on the sediments span-ning
the lower part of the post-evaporitic continen-tal
unit. Best suited outcrops of these sediments lie
close to the southeastern margin of the Fortuna basin,
where folding and tilting propitiate the exposure of
thick sediment successions. Two sections (Chorrico
and Sifo´n de Librilla) with numerous vertebrate fos-sil
traces were selected while taking into account the
need of a precise biochronologic constraint for this
study.
The Chorrico (CH) section is situated in the
Molina de Segura suburbs, 6 km to the north of
Murcia (Fig. 1), where beds dip 50º towards the
northwest. The base of the section directly overlies
the gypsiferous unit by means of a lateral-slip fault
contact. The CH section is about 350 m thick and
can be divided into two interfingering facies assem-blage
units (Fig. 2). The main unit consists of al-ternating
gypsiferous marls, silts, and conglomeratic
sandstones arranged in a prograding alluvial fan–fan
delta sequence. Mesozoic and Tertiary limestones
from the External Betics (western Fortuna margin)
are the main source rocks for these sediments. Pa-.72 M. Garce´s et al. / Earth and Planetary Science Letters 163 (1998) 69–81
Fig. 2. The Chorrico magnetostratigraphy. Legend: 1: sandstones
and conglomerates; 2: red silts; 3: light coloured silts; 4:grey
marls; 5: Red clays and breccias (‘Conglomerates of Murcia’
unit); 6: dark organic-rich silts; 8: mammal sites (in black
Paraethomys bearing sites).
leosols or other evidences of subaerial exposure are
not observed, but the abundance of levels with small
mammal fossils suggests a dominantly emergent or
very shallow transitional deltaic environment [6]. In
the studied section marine fossils, other than re-worked
planktonics from the underlying Tortonian
marls, are unknown. However, marine influxes in the
lowermost part cannot be totally excluded because
marine fossils are reported from laterally equivalent
(towards the NE) sediments [5].
In the upper part of the CH section, a second
facies assemblage interfingers with the main unit. It
consists of alluvial red clays and massive monogenic
breccias. Their immature sedimentary fabric denotes
the predominance of mass-flow processes and a rel-atively
short transport from the source area. In con-trast
to the CH main unit, the source rocks of these
sediments are Permotriassic phyllites and quartzites
corresponding to the Internal Betics, which crop out
in the massifs to the east of the Alhama de Murcia
Fault. This unit is referred to part of the so called
‘Conglomerates of Murcia’ [1], a marginal alluvial
unit associated with the erosion of the paleomassif
that extended to the east of the Alhama de Murcia
Fault since the late Tortonian.
The Sifo ´n de Librilla (SL) section [18] is exposed
along the eastern flank of a syncline structure to the
north of Librilla (Fig. 1). The sampled succession
shows a gentle dip (25 to 30º) towards the west.
The section consists of 360 m of alluvial red beds,
where red silts alternate with palustrine grey marls
and minor channelled conglomerates and sandstones
(Fig. 3). The fine-grained facies in the upper part of
the section shows a distinct alternation of alluvial red
silts and palustrine grey marls. Transition from allu-vial
to palustrine facies occurs through red-and-grey
mottled facies showing traces of hydromorphous pa-leosols.
The top of the grey marl intervals contains
organic-rich reduced facies, with abundant gastropod
shells and small mammal fossil remains. Very thin
laminated lacustrine limestones also occur associated
to the palustrine facies.
The SL section records a change in the sedimen-tary
conditions towards the top of the sequence. It
is marked by the retreat of the grey reduced facies,
providing evidence for a drastic decline of ponded
areas and the enhancement of the basin drainage.
Hydromorphous paleosols disappear and red allu-.M. Garce´s et al. / Earth and Planetary Science Letters 163 (1998) 69–81 73
Fig. 3. The Sifo´n de Librilla magnetostratigraphy. Closed circles
represent component A (primary) paleomagnetic directions, open
circles represent (secondary) component B directions. Legend:
1: sandstones and conglomerates; 2: red silts; 3: mottled silts
(hydromorphous paleosols); 4: grey marls; 5: limestones; 6:
carbonate crusts (caliche paleosols); 7: dark organic-rich silts; 8:
mammal sites (black portrays Paraethomys bearing sites).
vial sedimentation alternates with carbonate crusts
(caliche paleosols). The top of the section is formed
by a succession of thickly bedded conglomerates.
The conglomerate clasts mainly consist of Meso-zoic
and Tertiary limestones, indicating a main ter-rigenous
sediment supply from the External Betics
(western basin margin).
4. Mammal biostratigraphy
The Turolian continental stage [23] comprises
the European Mammal Neogene (MN) reference as-semblages
MN11 to MN13 [24,25]. In the eastern
Betics, a number of studies have contributed to con-struct
a high-resolution biostratigraphy for this time
slice [1,12–15]. Datable volcanic rocks, interbedded
within the sediments [5,11] and the lateral interfin-gering
between marine and continental units have
provided the mammal biozonation with some use-ful
constraints on absolute ages. In this region, the
late Turolian (MN13) faunas broadly correlate to
the Messinian [15,16]. Also, in the Cabriel Valley
(Valencia, east Spain), the MN12=MN13 boundary
has been approximately dated to 6.7 Ma (lower
Messinian) by means of magnetostratigraphy [26].
The MN13=MN14 (Turolian=Ruscinian) boundary,
traditionally positioned at the Messinian=Zanclean
boundary, is correlated in the Cabriel section to the
interval between chrons C3n.3n and C3n.4n (lower
Pliocene), with an age of 4.9 Ma [26].
The late Turolian mammal fauna in the Iberian
peninsula is characterised by the entry of a signif-icant
number of immigrants of African and Asi-atic
affinities ([13]). Among rodents, Stephanomys
and Apodemus are the most characteristic late Tur-olian
murids in the western European basins [27–
29]. An especially relevant bioevent is the en-try
of Paraethomys miocaenicus, an Asiatic immi-grant
which colonised both shores of the Mediter-ranean
basin [28–31]. The best-documented late
Turolian localities of southern Spain, containing
Paraethomys miocaenicus, are the classic sites of
Librilla [11,15,32,33], Crevillente 6 [15] and La Al-berca
[11,16]. The close similarities to the lower
Ruscinian (MN14) site of Alcoy (Alicante) sug-gest
that the fauna of La Alberca (at the northern
slopes of the Carrascoy massif, NE of Librilla) is.74 M. Garce´s et al. / Earth and Planetary Science Letters 163 (1998) 69–81
slightly younger than Crevillente 6 and Librilla, and
biochronologically close to the Turolian=Ruscinian
boundary [16]. The stratigraphic position of both La
Alberca and Crevillente 6, interbedded within a ma-rine
sequence with G. conomiozea [15,16], provides
assignment of Messinian Stage to these sites.
The biostratigraphic data of this study constrain
the stratigraphic datum for the entry of Paraethomys.
In the CH section (Fig. 2) sites CH5a, CH4a,
CH3b, and AU2a to AU2r yielded a character-istic
late Turolian assemblage with Stephanomys
ramblensis, Occitanomys adroveri, Apodemus gu-drunae,
Ruscinomys lasallei and Cricetus kormosi.
At sites AU3, Paraethomys miocaenicus occurs, to-gether
with Stephanomys ramblensis, Cricetus kor-mosi
and Ruscinomys lasallei.
In the SL section (Fig. 3), sites SIF-1, SIF-3,
and SIF-52 yield an assemblage with Apodemus aff.
gudrunae and Stephanomys ramblensis [14]. Occur-rence
of Paraethomys miocaenicus is found at sites
SIF-61 and SIF-79. The classic Paraethomys-bearing
site referred in the literature to as Librilla [11,14] can
be correlated to a level between SIF-61 and SIF-79.
In the same stratigraphic position, the occurrence
of camels assigned to the genus Paracamel us was
reported, which is the oldest record of this eastern
immigrant taxon in the Old World [32,33], together
with the site of Venta del Moro in Valencia [34].
The entry of eastern immigrants such as Para-camelus
aguirrei in southern Spain during the late
Miocene is suggested to proceed via North Africa
[35] since this species is unknown in the rest of
Fig. 4. NRM demagnetization diagrams of representative samples from the Chorrico section. All the projections in tectonic corrected
coordinates.
the European Turolian large mammal sites. Among
rodents, the entry of Paraethomys may represent the
same dispersal event since the first appearance of
this species occurs in Librilla at the same strati-graphic
position as Paracamel us. Further evidences
of a late Turolian intercontinental faunal exchange
are the occurrences of some rodent genera of west-ern
European origin such as Apodemus, Prolagus,
Occitanomys and Cricetus in the Ai ¨t Kandoula
Basin (High Atlas, Morocco) [36]. This implies an
ephemeral land bridge between north Africa and
western Europe prior to the late Messinian isolation
of the Mediterranean basin [37].
5. Paleomagnetic analysis
The paleomagnetic study of the SL and CH sec-tions
is based on the analysis of 300 magnetostrati-graphic
sites, cored in the field with a portable
drill. The natural remanent magnetisation (NRM)
was studied in the laboratory by means of standard
stepwise thermal (TH) demagnetisation procedures.
Additional IRM experiments were carried out to
check for differences on the ferromagnetic composi-tion
among lithologies and to identify the remanence
carriers.
5.1. The Chorrico (CH) section
Thermal demagnetisation of samples from the
CH section reveals a two-component NRM (Fig. 4)..M. Garce´s et al. / Earth and Planetary Science Letters 163 (1998) 69–81 75
Tabl e 1
Mean directions and Fisher statistics of the Chorrico and Sifo ´n de Librilla sections
Sites N Geographic coordinates Bedding coordinates
dec inc k a95 dec inc k a95
Chorrico
Normal 68 278 40 12 5 326 30 12 5
Reversed 52 96 23 10 6 120 22 11 6
Sifo´n de Librilla
component A Normal 59 26 67 19 4 349 41 17 5
Reversed 36 221 59 8 9 189 37 8 9
component B Reversed 44 219 70 10 7 175 47 12 7
The low temperature (up to 300ºC) component in
geographic coordinates yields an average direction
which approximately conforms to the present day
field dipole. A characteristic remanent magnetisation
(ChRM) is isolated above 300ºC and shows both
normal and reversed polarities. Thermal decay of the
ChRM frequently shows steep intensity decreases
below 590ºC, but higher temperatures (up to 680ºC)
are required for complete demagnetisation of most
of the samples. This suggests that NRM is carried by
iron oxides such as hematite and magnetite. Normal
and reversed ChRM directions do not show perfect
antipodal directions (angle between normal and re-verse
polarity sets equals to 17º), and fail to pass the
reversal test [38], suggesting an incomplete cleaning
of secondary components. The overall mean direc-tion
(reverse samples rotated to antipodal) yields
309=27 after correction for bedding tilt, thus report-Fig.
5. Stereonet projection of paleomagnetic directions from the CH and SL sections after correction for bedding tilt. See Fisher
statistics in Table 1.
ing an apparent vertical axis rotation of as much
as 50º (Table 1, Fig. 5). The paleomagnetic data
indicates a very substantial anticlockwise rotation
in this sector, which can be linked to the prevalent
left-lateral shear associated to the Alhama de Murcia
wrench fault [22,39].
5.2. The Sifo´n de Librilla (SL) section
Thermal demagnetisation of samples from the SL
section frequently yields two paleomagnetic compo-nents
after removal of the secondary low temperature
magnetisation (Fig. 6). The unblocking temperature
spectra of these two components do not overlap
significantly and their directions can be properly
calculated by means of least-square analysis of the
demagnetisation data [40].
The highest temperature component (component.76 M. Garce´s et al. / Earth and Planetary Science Letters 163 (1998) 69–81.M. Garce´s et al. / Earth and Planetary Science Letters 163 (1998) 69–81 77
A) is invariably present in samples of red and mottled
silts (Fig. 6). Its maximum unblocking temperatures
are well above 660ºC and this reveals the presence of
hematite as the main remanence carrier. Component
A yields both normal and reverse polarities (Fig. 5),
recording a sequence of four stratigraphic polarity
zones (Fig. 3). Mean directions in geographic coor-dinates
yield reliable inclinations but declinations are
rotated clockwise. After correction for bedding tilt,
mean directions conform to a northerly (unrotated)
shallow magnetisation (Table 1).
In addition to component A, an intermediate tem-perature
paleomagnetic magnetisation (component
B) is defined in the range 250–500ºC. It is reg-ularly
present in all samples from the grey marl
intervals, but also in many red-grey mottled sam-ples
and in some red silts close to the greyish fa-cies.
Component B always yields reverse polarity
(Fig. 5). The mean direction in geographic coordi-nates
averages 219= 70, and 175= 47 after bedding
correction, thus supporting an early pre-tilt magneti-sation
(Fig. 5). Both NRM unblocking temperatures
and coercivity spectra from IRM acquisition curves
suggest the presence of magnetite as the principal
magnetic phase in the grey sediments, and the most
likely carrier of the component B (Fig. 6).
The fact that component A records a consistent
sequence of polarity reversals (Fig. 3) supports an
early (primary) magnetisation. In contrast to the re-sults
from the CH section, the mean paleomagnetic
direction in the SL section does not support vertical
axis rotations in this sector. Magnetisation of com-ponent
A in the SL section was likely affected by
appreciable inclination shallowing upon deposition
and=or subsequent compaction, similar to the CH
section.
We must conclude that component B represents an
early (pre-tilt), but secondary magnetisation, which
selectively affected the organic-rich greyish facies.
The fact that component B is tightly clustered to a
reversed paleomagnetic direction suggests that the
Fig. 6. NRM demagnetization diagrams of samples from a particular stratigraphic interval in the Sifo´n de Librilla section with cyclic
alternations of red alluvial silts and palustrine=lacustrine grey marls. A high temperature component (A) is present in all the red samples.
In some of these, an intermediate temperature component (B) is also recognised (sites 64 and 80). Component B is prevailing in the grey
intervals (site 67), where component A may be absent. Component B always yields reverse polarity, while component A records up to
three polarity transitions (see Fig. 3). All the projections in tectonic corrected coordinates.
demagnetisation is associated to a single discrete
(short-lived) event, rather than to a process operat-ing
continuously through diagenesis. This particular
event must have taken place within a period of re-verse
geomagnetic field.
6. Magnetostratigraphic correlation and
biochronological implications
The virtual geomagnetic pole (VGP) latitudes of
the characteristic paleomagnetic directions yield a
well defined magnetic polarity sequence in both the
CH and SL sections (Figs. 2 and 3). In the CH
VGP latitudes were calculated after correcting the
paleomagnetic declinations for the observed vertical
axis rotation discussed above. The appearance of
Paraethomys in both sections allows an unequivocal
correlation between the two magnetic polarity se-quences
(Fig. 7). Given the additional constraint that
the mammal assemblages found in the sections are of
Messinian age [15,16], a unique correlation with the
geomagnetic polarity time scale is feasible (Fig. 7).
The resulting absolute chronology implies that the
composite sequence of the continental Rambla Sal-ada
Formation in the Fortuna basin represents more
than 1 My, covering the early to latest Messinian,
from approximately 6.8 to 5.7 Ma. The entry of
the eastern immigrants Paracamelus aguirrei and
Paraethomys miocaenicus is accurately dated in the
lower part of chron C3An.1n at an age of approxi-mately
6.1 Ma.
Our correlation is in good agreement with the data
from the Cabriel basin (eastern Spain), where the
late Turolian fauna of Venta del Moro is correlated
to lower chron C3r [26]. In this concern, the age
of Venta del Moro is of special value because its
fauna also includes Paraethomys miocaenicus and
Paracamelus aguirrei and shares most of the other
elements with the Paraethomys-bearing localities of
our sections. The entry of eastern immigrants in.78 M. Garce´s et al. / Earth and Planetary Science Letters 163 (1998) 69–81
Fig. 7. Correlation of the Messinian alluvial and lacustrine sequences in the Fortuna basin with the geomagnetic polarity time scale
and the marine and continental chronologies. The age of Samos 5 [11], Librilla (this study) and Venta del Moro [26] are based on
magnetostratigraphy. The age of Crevillente 6 and La Alberca are based on marine–continental correlations.
Spain is in full coincidence with the reported age
for the arrival of European immigrants into northern
Africa [37]. This further supports a late Messinian
intercontinental mammal exchange starting at least
as early as 6.1 Ma, well before the main evaporitic
phase of the Mediterranean salinity crisis which is
magnetostratigraphically dated at 5.8 Ma [41].
Earlier magnetostratigraphic datings in the
Cabriel valley [26] yielded an approximate age for
the not well-constrained MN12=MN13 boundary of
6.7 Ma, below the base of C3An. In the CH sec-tion,
the earliest late Turolian (MN13) mammal site
occurs somewhat older in chron C3Ar, at approx-imately
6.8 Ma. Thus, from this study, an age for
the MN12=MN13 boundary older than 6.8 Ma is
preferred. This contradicts earlier correlations of
the MN12 reference mammal site of the Casa del
Acero to the late Messinian, which were inferred
from its stratigraphic location within the evaporitic
sequences in the Fortuna basin [12]. However, as
discussed below, the supposed late Messinian age of
the evaporites in the Fortuna basin is untenable and,
in consequence, the precise age of Casa del Acero is
unresolved.
Our ages are congruent with the magnetostratig-raphy
of Samos (Greece) where basal MN 13 fauna
is correlated to chron C3Ar [42], and with the re-cently
reported 40 Ar= 39 Ar datings of mammal faunas
from eastern Mediterranean and Iran [43] suggesting
a correlation of the base of MN13 to the base of the
Messinian stage [44].
7. Chronology of the Fortuna basin and
consequences for the Messinian evolution of the
western Mediterranean
The magnetostratigraphic dating of the post-evap-oritic
alluvial sequences provides new clues for the
chronology of the basin infill. In the CH section,.M. Garce´s et al. / Earth and Planetary Science Letters 163 (1998) 69–81 79
sediments record the emergence of a confined shal-low
basin during the early Messinian (older than
6.8 Ma). The overall sequence represents a progra-dation
of the conglomeratic alluvial fans fed from
the northwestern basin margin. These interfinger
with the Conglomerates of Murcia, which consist
of smaller scale alluvial fans fed from the opposing
southeastern margin. In the SL section we observe
that alluvial and palustrine sedimentation alternated
for almost 1 My, from middle to late Messinian
(at least between 6.5 and 5.8 Ma). The frequency
of reducing palustrine facies indicates an inefficient
basin drainage, which prompted the development of
ephemeral ponded areas. The scenario changes at
the top of the section, where palustrine deposits dis-appear
and an overall progradation of the alluvial
system takes place. Hydromorphous paleosols are
replaced by caliche crusts, suggesting a drier and
better drained environment. This change occurs in
the lower part of chron C3r at approximately 5.8 Ma,
and is most likely coeval to the decrease of the base
level associated with the beginning of the Messinian
salinity crisis.
We conclude that basin confinement and the fol-lowing
precipitation of evaporites in the Fortuna
basin has taken place earlier, during the latest Torto-nian
to the earliest Messinian, indicating that these
evaporites are not coeval with, but predated the evap-orites
of the Mediterranean salinity crisis. Therefore,
the inclusion of the Fortuna basin in a hypothetical
Betic marine corridor during the late Messinian [4]
is untenable. The primary role of the eustatism in
the Fortuna basin evolution [6] was criticised earlier
[5] because it ignored the tectonic deformation that
took place during the deposition of the Miocene se-quences
in the Fortuna basin. Lukowski [5] stressed
the importance of the tectonic uplift of the basin mar-gins
as the cause for the isolation of the basin and the
precipitation of the evaporites. The early initiation of
confinement leading to marine regression and ter-restrial
environments is not restricted to the Fortuna
basin. Other intramontane basins, lying on the con-tact
between the External and Internal Betics, show
a comparable evolution. The age of the emergence
of the Granada [45,46] and Guadix–Baza basins [47]
which resulted from the orogenic uplift coupled with
a period of eustatic downfall, is correlated to late
Tortonian–early Messinian.
8. Remagnetisation in the Sifo ´n de Librilla
section
In the SL section, remagnetisation typically oc-curs
in the grey intervals (component B in Fig. 3).
Although a detailed analysis of its origin is beyond
the scope of this paper, this subject deserves a de-tailed
separate study because it is of fundamental
interest and has important implications for the ac-quisition
of the NRM in this type of (palustrine–
alluvial) sediments.
From what we have observed the remagnetisation
is likely related to a discrete event during a period
of reversed polarity which caused an intermediate
temperature reversed component to originate in all
grey beds of the section. The remagnetisation must
have taken place in an early stage because of its
pre-folding age. A likely period for this event is the
upper reverse magnetozone in the Librilla section,
where we find the change from a palustrine–alluvial
to a prograding alluvial system, which is suggested
linked to the base level drop of the Mediterranean
late Messinian. A possible cause for the remagnetisa-tion
component is that this base level drop caused an
increased and efficient drainage of groundwater, thus
exposing a reduced environment (the grey layers) to
oxidation. Oxidation of iron sulphides to magnetite
would cause a newly formed chemical remanent
magnetisation, acquiring the then ambient (reversed)
field. The suggested driving mechanism for fluid mi-gration
in the SL section, would be a eustatic base
level lowering, in contrast to other reported ancient
remagnetisations where fluid migration is induced by
tectonic stress [48].
9. Conclusions
In the Fortuna basin, the Messinian Stage is
mainly represented by continental alluvial-fan to fan-delta
and lacustrine sediments of the Rambla Salada
Formation, overlying the shallow marine diatomitic–
evaporitic sequences usually referred to the Chicamo
Formation [4]. The initiation of the terrestrial envi-ronments
in the Fortuna basin correlates to a latest
Tortonian–earliest Messinian event, similar to other
intramontane basins of the central Betics.
The oldest fossil mammal locality in the studied.80 M. Garce´s et al. / Earth and Planetary Science Letters 163 (1998) 69–81
sections is attributed to MN 13 and is correlated
to chron C3Ar at an age of 6.75 Ma. Thus, from
this study an age for the MN12=MN13 boundary
older than 6.8 Ma is preferred. The emergence of the
Betics intramontane basins is followed by the first
terrestrial faunal exchange between Africa and Eu-rope
at 6.1 Ma, when western European rodents ar-rive
in northern Africa [37,36] and genera of Asiatic
affinities appear in the Fortuna Basin (this study).
Consequently, the tectonic uplift leading to the emer-gence
of the central and eastern Betic basins also
disrupted the seaway between the Mediterranean and
the Atlantic, prompting intercontinental land-mam-mal
exchange.
The proposed chronology revises previous mod-els,
in which the evaporitic phases in the Fortuna
basin were correlated to the Messinian salinity crisis,
and in which eustatic sea-level change was suggested
the primary controlling factor in the progressive iso-lation
of this marginal basin from the open sea
[4,6,8]. Also, the inclusion of the Fortuna basin in
a hypothetical Betic marine corridor during the late
Messinian [4] is no longer tenable.
Acknowledgements
Supported by the Netherlands Geosciences Foun-dation
(GOA) with financial aid from the Nether-lands
Organisation of Scientific Research (NWO),
the spanish Ministry of Education and Science PB94-
1265 project, the Comissionat per Universitats i Re-cerca
de la Generalitat de Catalunya Grup de Quali-tat
GRQ94-104, and EC Project CI1-CT94-0114. We
thank M. Freudenthal, E. Marti ´n-Sua´rez, M. Llenas
and M. Guitart for their collaboration in the bios-tratigraphic
samplings. We also thank C. Langereis,
M. Woodburne and E. Lindsay for reviewing the
manuscript. [RV]
References
[1] C. Montenat, Les formations ne´oge`nes et quaternaires du
Levant espagnol (Provinces d’Alicante et de Murcia), Ph.D.
Thesis, Universite ´ d’Orsay, 1973, 1170 pp.
[2] C. Montenat, Le Miocene terminal des chaines betiques
(Espagne meridionale). Esquisse paleogeographique, Proc.
Kon. Ned. Akad. Wetensch. 180–187, 1973.
[3] C. Sanz de Galdeano, J.A. Vera, Stratigraphic record and
palaeogeographical context of the Neogene basins in the
Betic Cordillera, Spain, Basin Res. 4 (1992) 21–36.
[4] D.W. Mu¨ller, K.J. Hsu¨ , Event stratigraphy and paleoceanog-raphy
in the Fortuna basin (Southeast Spain): A scenario
for the Messinian salinity crisis, Paleoceanography 2 (1987)
679–696.
[5] P. Lukowski, Evolution tectonosedimentaire du bassin Neo-gene
de Fortuna (Cordille `res be´tiques orientales, Espagne),
Ph.D. Thesis, Universite´ de Paris Sud, 1987, 403 pp.
[6] C. Santisteban, Petrologi´a y sedimentologi´a de los ma-teriales
del Mioceno superior de la cuenca de Fortuna
(Murcia), a la luz de la ‘teori ´a de la crisis de salinidad’,
Universitat de Barcelona, 1981, 722 pp.
[7] F. Orti ´, J. Garci´a-Veigas, L. Rosell, J.M. Rouchy, M. Ingle`s,
D. Gimeno, A. Kasprzyk, E. Playa`, Correlacio´n litoestrati-gra´fica
de las evaporitas messinienses en las cuencas de
Lorca y Fortuna (Murcia), Geogaceta 14 (1993) 98–101.
[8] D.W. Mu¨ller, H. Schrader, Diatoms of the Fortuna basin,
Southeast Spain: evidence for the intra-Messinian inunda-tion,
Paleoceanography 4 (1989) 75–86.
[9] P. Lukowski, R. Wernli, A. Poisson, Mise en e´vidence de
l’importance des de´poˆts messiniens dans le bassin Miocene
de Fortuna (Province de Murcia, Espagne), C.R. Acad. Sci.
Paris 307 (1988) 941–947.
[10] F.J. Sierro, J.A. Flores, J. Civis, I. Zamarren˜o, A. Va´zquez,
C. Santisteban, J. Porta, Las margas de Fortuna: Bioestrati-grafi
´a y caracterizacio ´ n paleoceanogra´fica, III Congr. Geol.
Espan˜a, 1992, pp. 222–226.
[11] C. Montenat, L. Thaler, J.A. Van Couvering, La faune
de rongeurs de Librilla. Corre´lations avec les forma-tions
marines du Mioce`ne terminal et les datations ra-diometriques
du volcanisme de Barqueros (Province de
Murcia, Espagne me´ridional), C.R. Acad. Sci. Paris 281
(1975) 519–522.
[12] J. Agusti ´, S. Moya`-Sola`, J. Gibert, J. Guille ´n,M.Labrador,
Nuevos datos sobre la bioestratigrafi´a del Neo´ geno conti-nental
de Murcia, Paleontol. Evol. 18 (1985) 83–93.
[13] P. Mein, J. Agusti ´, Les Gisements de Mammife`res neogenes
de la zone Betique, Doc. Trav. IGAL 13 (1990) 81–84.
[14] J. Agusti ´, M. Freudenthal, J.I. Lacomba, E. Marti ´n Sua´rez,
Mio–Pliocene transition in the Librilla section (Mula basin,
Murcia, Spain), Global events and Neogene evolution of the
Mediterranean, IX R.C.M.N.S. Congr., 1990, 15 pp.
[15] H. De Bruijn, P. Mein, C. Montenat, A. Van de Weerd, Cor-re´lations
entre les gisements de rongeurs et les formations
marines du Mioce`ne terminal d’Espagne me´ridionale, Proc.
Kon. Ned. Akad. Wetensch. 78 (1975) 1–32.
[16] P. Mein, G. Bizon, J.J. Bizon, C. Montenat, Le gisement de
Mammife`res de La Alberca (Murcia, Espagne me´ridionale).
Corre ´lations avec les formations marines du Mioce`ne termi-nal,
C.R. Acad. Sci. Paris, Se´rie D 276 (1973) 3077–3080.
[17] C. Sanz de Galdeano, Geologic evolution of the Betic
Cordilleras in the Western Mediterranean Miocene to the
present, Tectonophysics 172 (1990) 107–119.
[18] C. Montenat, P. Ott d’Estevou, J. Rodri ´guez-Ferna´ndez,
C. Sanz de Galdeano, Geodynamic evolution of the Betic.M. Garce´s et al. / Earth and Planetary Science Letters 163 (1998) 69–81 81
Neogene intramontane basins (S and SE Spain), Paleontol.
Evol. 2 (1990) 5–59.
[19] C. Montenat, P. Ott d’Estevou, Eastern Betic Neogene
basins — A review, Doc. Trav. IGAL 13 (1990) 9–15.
[20] C. Sanz de Galdeano, Los accidentes y fracturas principales
de las Cordilleras Be´ticas, Est. Geol. 39 (1983) 157–165.
[21] F. Gauyau, R. Bayer, J.C. Bousquet, J.C. Lachaud, A.
Lesquer, C. Montenat, Le prolongement de l’accident
d’Alhama de Murcia entre Murcia et Alicante (Espagne
me´ridionale). Resultats d’une e´tude ge´ophysique, Bull. Soc.
Ge´ol. Fr. 19 (1977) 623–629.
[22] P.G. Silva, D.L. Goy, C. Zazo, Caracteri´sticas estructurales
y geome´tricas de la falla de desgarre de Lorca–Alhama,
Geogaceta 12 (1992) 7–10.
[23] M. Crusafont, Observations a` un travail de M. Freudenthal
et P.Y. Sondaar sur des nouveaux gisements a` Hipparion
d’Espagne, Proc. Kon. Ned. Akad. Wetensch. B 68 (1965)
121–126.
[24] V. Fahlbusch, Report on the international symposium on
mammalian stratigraphy of the European Tertiary, Newsl.
Stratigr. 5 (1976) 160–167.
[25] H. De Bruijn, R. Daams, G. Daxner-Ho¨ck, V. Fahlbusch,
L. Ginsburg, P. Mein, J. Morales, Report of the RCMNS
working group on fossil mammals, Reisenburg 1990,
Newsl. Stratigr. 26 (1992) 65–118.
[26] N.D. Opdyke, P. Mein, E. Lindsay, A. Perez-Gonza´lez,
E. Moissenet, V.L. Norton, Continental deposits, magne-tostratigraphy
and vertebrate paleontology, late Neogene
of Eastern Spain, Palaeogeogr. Palaeoclimatol. Palaeoecol.
133 (1997) 129–148.
[27] A. van de Weerd, Rodent faunas of the Mio–Pliocene con-tinental
sediments of the Teruel–Alfambra region, Spain,
Utrecht Micropaleontol. Bull Spec. Publ. 2, 1976, 217 pp.
[28] P. Mein, E. Moissenet, R. Adrover, Biostratigraphie du
Ne´oge`ne Supe´rieur du bassin de Teruel, Paleontol. Evol. 23
(1990) 121–139.
[29] J. Agusti ´, M. Llenas, The late Turolian muroid rodent
succession in eastern Spain, Acta Zool. Cracov. 39 (1996)
47–56.
[30] J.J. Jaeger, N. Lo´pez Marti ´nez, J. Michaux, L. Thaler,
Les faunes de micromammife`s du Ne´oge`ne superieur de
la Mediterrane´e occidentale. Biochronologie, corre´lations
avec les formations marines et e´changes intercontinentaux,
Bull. Soc. Ge´ol. Fr. 7 (1977) 501–506.
[31] H. Thomas, R.L. Bernor, J.J. Jaeger, Origines du peuple-ment
mammalien en Afrique du Nord durant le Mioce`ne
terminal, Geobios 15 (1982) 283–297.
[32] C. Montenat, M. Crusafont, De´couverte de mammife`res
dans le Ne´oge`ne et le Ple´istoce`ne du levant espagnol
(provinces d’Alicante et de Murcia), C.R. Acad. Sci. Paris
270 (1970) 3194–3197.
[33] M.T. Alberdi, J. Morales, S. Moya`-Sola`, B. Sanchiz,
Macrovertebrados (reptilia y mammalia) del yacimiento fini-mioceno
de Librilla (Murcia), Est. Geol. 37 (1981) 307–312.
[34] J. Morales, D. Soria, E. Aguirre, Came´lido finimioceno en
Venta del Moro Primera cita para Europa occidental, Est.
Geol. 36 (1980) 139–142.
[35] M. Pickford, J. Morales, D. Soria, First fossil camels from
Europe, Nature 365 (1993) 701.
[36] M. Benammi, B. Orth, M. Vianey-Liaud, Y. Chaimanee,
V. Suteethorn, G. Feraud, J. Hernandez, J.J. Jaeger, Mi-cromammife`res
et biochronologie des formations ne´oge`nes
du flanc sud du Haut-Atlas Marocain: implications bioge´o-graphiques,
stratigraphiques et tectoniques, Afr. Geosci.
Rev. 2 (1995) 279–310.
[37] M. Benammi, M. Calvo, M. Pre´vot, J.J. Jaeger, Magne-tostratigraphy
and paleontology of Ai¨t Kandoula basin
(High Atlas, Morocco) and the African–European late
Miocene terrestrial fauna exchanges, Earth Planet. Sci.
Lett. 145 (1996) 15–29.
[38] P.L. McFadden, M.W. McElhinny, Classification of the
reversal test in paleomagnetism, Geophys. J. Int. 103
(1990) 725–729.
[39] P. Lukowski, A. Poisson, Le bassin de Fortuna, Doc. Trav.
IGAL 13 (1990) 303–311.
[40] J.L. Kirschvink, The least-squares line and plane and the
analysis of paleomagnetic data, Geophys. J. R. Astron. Soc.
62 (1980) 699–718.
[41] F. Gautier, G. Clauzon, J.P. Suc, J. Cravatte, D. Violanti,
Age et dure´e de la crise de salinite ´ messinienne, C. R.
Acad. Sci. Paris 318 (1994) 1103–1109.
[42] S. Sen, J.P. Valet, Magnetostratigraphy of late Miocene
continental deposits in Samos, Greece, Earth Planet. Sci.
Lett. 80 (1986) 167–174.
[43] R.L. Bernor, N. Solounias, C.C.I. Swisher, J.A. van
Couvering, The correlation of three classical ‘Pikermian’
mammal faunas — Maragheh, Samos and Pikermi —
with the European MN unit system, in: R.L. Bernor,
V. Fahlbusch, H.W. Mittmann (Eds.), The Evolution of
Western Eurasian Neogene Mammal Faunas, Columbia
University Press, New York, 1996, pp. 137–154.
[44] F.F. Steininger, W.B. Berggren, D.V. Kent, R.L. Bernor, S.
Sen, J. Agusti ´, Circum Mediterranean Neogene (Miocene
and Pliocene) marine–continental chronologic correlations
of European mammal units and zones, in: R.L. Bernor,
V. Fahlbusch, H.W. Mittmann (Eds.), The Evolution of
Western Eurasian Neogene Mammal Faunas, Columbia
Press, New York, 1996, pp. 7–46.
[45] J. Rodri ´guez-Ferna´ndez, J. Ferna´ndez, A.C. Lopez Garrido,
C. Sanz de Galdeano, The central sector of the Betic
Cordilleras, a realm situated between the Atlantic and
Mediterranean domains during the upper Miocene, Ann.
Ge´ol. Pays Hell. 32 (1984) 97–103.
[46] J. Rodri ´guez-Ferna´ndez, C. Sanz de Galdeano, Onshore
Neogene Stratigraphy in the North of the Alboran Sea
(Betic Internal Zones): Paleogeographic implications,
Geo-Mar. Lett. 12 (1992) 123–128.
[47] J.M. Soria, Evolucio ´n sedimentaria y paleogeogra´fica
durante el Mioceno superior en el borde norte de la cuenca
de Guadix, Cordillera Be´tica Central, Est. Geol. 50 (1994)
59–69.
[48] D. Suk, D.R. Peacor, R. Van der Voo, Replacement of pyrite
framboids by magnetite in limestone and implications for
palaeomagnetism, Nature 345 (1990) 611–613.