gibert et al. (2007) - the early to middle pleistocene boundary in the baza basin (spain)

8/7/2019 GIBERT et al. (2007) - The Early to Middle Pleistocene boundary in the Baza Basin (Spain)

1/23

Quaternary Science Reviews 26 (2007) 20672089

The Early to Middle Pleistocene boundary in the Baza Basin (Spain)

L. Giberta,, G. Scottb, R. Martinc, J. Gibertd

aDepartment Enginyeria Minera i Recursos Naturals, Universitat Politecnica de Catalunya, Av. Bases de Manresa 61-73, Manresa 08240, SpainbBerkeley Geochronology Center, 2455 Ridge Road, Berkeley, CA 94709, USA

cDepartment of Biological Sciences, Murray State University, Murray, KY 42071, USAdInstitut de Paleontolog a M.Crusafont, Escola Industrial 23, Sabadell 08201, Spain

Received 31 May 2006; received in revised form 2 May 2007; accepted 17 June 2007

Abstract

The continental vertebrate fauna of Cullar Baza-1 (Granada, Spain) occur immediately above the Matuyama/Brunhes polarity

boundary, and therefore represent an initial record for the Middle Pleistocene. All polarity zones for the Late Pliocene through Middle

Pleistocene are found in an 80 m section, which includes this fossil quarry. The existing collection of abundant remains of micro-

mammals, macro-mammals and lithic artifacts (indicating human activity) can now be assigned an earliest Middle Pleistocene age of

0.75 Ma. Another section, 25 km across the Neogene Baza Basin, also has the Matuyama/Brunhes polarity boundary. However, in this

case the Huescar-1 fossil quarry and the Puerto Lobo sites are both below the polarity boundary, thus representing a record of the late

Early Pleistocene at $0.9 Ma. Differences in micro-mammal species across these Matuyama/Brunhes boundaries are significant and

justify an approximately coincident biostratigraphic boundary between the Biharian and Toringian stages.

r 2007 Elsevier Ltd. All rights reserved.

1. Introduction

The intra-montane Guadix-Baza Basin is the largest of

the late Neogene inland basins of the Betic Cordillera. As a

consequence of the Alpine Orogeny, this Basin was isolated

from the sea during the late Miocene (Sanz de Galdeano

and Vera, 1992). The Basin had an internal drainage of

approximately 4000 km2 and could be divided into two

sub-basins. Generally, the Guadix Basin to the southwest

was filled with alluvial deposits, while lacustrine deposits,

along with fluvial/alluvial marginal deposits (Fig. 1),

dominated the Baza Basin to the northeast.

Six hundred meters of late Neogene continental sedi-

mentary infilling is exposed in the marginal northeasternpart of the Baza Basin (Gibert L, 2006). Gravity and

seismic studies indicate 42500 m of lacustrine sediments

buried near the town of Baza (Alfaro et al., in press). The

paleo-environments within the Baza Basin included a

shallow saline lake complex, surrounded by saline mud-

flats, freshwater ponds, small rivers with floodplains and

deltas (Gibert L. et al., 2005, 2007). The grass-covered

upland slopes merged into alluvial fans from the surround-

ing mountains. The Baza Basin continued accumulatingsediment until tectonic movements and stream piracy

initiated rapid and prolonged erosion (Calvache and

Viseras, 1997), stripping the softer, upper strata and

leaving behind the more resistant layers that make up the

present-day topography.

The Baza Basin has produced 440 fossil vertebrate sites

of Late Miocene, Pliocene and Pleistocene age, along

with some lithic artifact sites of Pleistocene age showing the

activity of early man (Ruiz-Bustos, 1976, 1984; Gibert J.

et al., 1998, 2006). Magnetostratigraphic studies from the

northeastern part of the Basin (Gibert L. et al., 2006;

Scott et al., 2007) have generated a polarity sequencefor the Orce-Venta Micena sections (three magnetozones,

reversenormalreverse) revealing the Pliocene/Pleistocene

boundary and calibration of the MN17 (mammal

Neogene zone) upper boundary. The youngest deposits

in these sections are of Early Pleistocene age (pre-Jaramillo

chron or 41.1 Ma), bounded by an upper erosion

surface at 965 m asl. Other areas in the Basin have supplied

fossils that indicate younger fauna than in the Orce-Venta

Micena area; these are Puerto Lobo-Huescar and Cullar,

the focus of this report. Both of these younger areas have

ARTICLE IN PRESS

0277-3791/$ - see front matter r 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.quascirev.2007.06.012

Corresponding author. Tel.: +34 937472423.

E-mail address: [email protected] (L. Gibert).
http://dx.doi.org/10.1016/j.quascirev.2007.06.012mailto:[email protected]:[email protected]://dx.doi.org/10.1016/j.quascirev.2007.06.012


2/23

well-exposed sedimentary sections, interbedded with fossi-

liferous sites and thus are well suited for detailedmagnetostratigraphic research. This study furthers our

general goal of developing a wider and more complete

chronologic and biostratigraphic framework for the Baza

Basin. All geologic observations in this report are the

product of the authors, unless specifically referred.

2. The Cullar section

In 1975, Cullar Baza-1 (CB-1) was the initial paleonto-

logical excavation in the Baza Basin, which now includes

numerous archeological/paleontological quarries, mostly

concentrated in the Orce area, 20 km to the NE (e.g. Venta

Micena, Fuente Nueva-3). An early Middle Pleistocene age

was assigned to CB-1 using faunal criteria (Ruiz-Bustos

and Michaux, 1976; Ruiz-Bustos, 1984).

A preliminary magnetostratigraphic and paleontologic

study was made $2 km NW from the CB-1 quarry (UTM

30537120E, 4159537N, Z 890905; Fig. 2) (Agust et al.,

1999). Situated $50 m stratigraphically below the CB-1

level, that study consisted of a 14m section with 10

paleomagnetic and three paleontologic levels. The Cullar

section in this current report is meant to supersede and

expand these preliminary observations. We will emphasize

the following advantages: (A) an additional 70 m of

overlying strata were sampled; (B) the paleontological

quarry CB-1 was included within the new section; (C) most

of the samples from the new collection have high-qualityremanent polarities (mostly reddish-brown siltstones and

paleosols); and (D) field tests were conducted to indicate

the stability and antiquity of remanence. By incorporating

the results from the preliminary section (Agust et al., 1999)

into this expanded section, the following features were

revealed: (1) the paleontological site CU-A will not be

considered further, since it has only a sparse collection of

post-cranial bones, as none were originally reported

(Agust et al., 1999); (2) the upper 4 m of that preliminary

section are not part of the Baza Formation, but are

unconsolidated or poorly consolidated claystones, sand-

stones and conglomerates composed of intra-basin clasts

and discontinuous calcrete beds that belong to the much

younger post-erosional, valley in-fill deposits; and (3) the

paleontological site CU-C, within these post-Baza deposits,

is Late Pleistocene or latest Middle Pleistocene, as are the

uppermost three normal polarity samples in the prelimi-

nary report. Our observations indicate that these young

post-Baza Formation deposits are extensively preserved

along the flanks of the present Cullar River Valley and are

significantly younger than the erosive opening of the Basin

(Fig. 3).

In the present study, a magnetostratigraphic section was

developed 2 km SE of the town of Cullar. This section

starts near the Cullar River (UTM coordinates:

ARTICLE IN PRESS

Fig. 1. Location maps. The Baza Basin (SE Iberian peninsula) showing the stratigraphic sections and paleontological sites. 1Cullar section and quarry

CB-1; 2Puerto Lobo sections and sites; 3Loma Quemada sites; 4Hue scar-3 site; 5Sabinar section; 6Orce sections and quarries FN-1, VM, BL-

5, and FN-3; 7Oria road section.

L. Gibert et al. / Quaternary Science Reviews 26 (2007) 206720892068


3/23

ARTICLE IN PRESS

Fig. 2. Aerial photos. The Cullar (top) and Puerto Lobo (bottom) areas, showing the primary and ancillary magnetostratigraphic sections as dashed lines.

L. Gibert et al. / Quaternary Science Reviews 26 (2007) 20672089 2069


4/23

30539063E, 4159055N, Z 900), follows the Oria road

and then the side valley towards the CB-1 quarry

(30538659E, 4158276N, Z 966) until the top of the

valley (30538709E, 4158186N, Z 975). The upper 5 m

of the section were collected in the next valley to the

west (30539182E, 4157813N, Z 985). The section

totals 81 m in thickness, starting above Mesozoic carbo-

nates, including the paleontological quarry CB-1 (Fig. 2),

and ending with the uppermost strata below the

erosive llano surface. This new section was made in

fresh outcrops or road cuts of alluvial fan and palustrine

facies. Paleomagnetic samples were taken from the

finer-grained units (primarily siltstones) between the

conglomerate beds and lenses. Conglomeratic channels

provided an opportunity to apply the conglomerate/tilt test

to fine-grained blocks swept from the banks of the paleo-

channels.

3. The Puerto Lobo section

The Puerto Lobo section was located in the northern

margin of the Baza Basin, 3 km SE of the town of Huescar

and 25 km north from the Cullar section (Fig. 1). This area

is adjacent to a far-ranging normal fault (Soria et al., 1987)

with a minimum estimated displacement of 30 m (direction

049/70N). In a previous study of a subaqueous landslide of

Early Pleistocene age, a few reverse polarity samples were

collected from the upthrown side of this fault (Gibert L. et

al., 2005). On the downthrown side of the fault are the

stratigraphically higher fossil sites of Puerto Lobo and

Huescar-1, along with the new magnetostratigraphic

sections (Fig. 2).

The Puerto Lobo section (UTM coordinates 30543408E,

4182540N, Z 935965; Fig. 2) starts in strata from the

distal fluvial facies (mudflats, overbank deposits) rich in

ARTICLE IN PRESS

Fig. 3. Photos of post-Baza Formation outcrops. Preliminary Cullar paleomagnetic section ofAgust et al. (1999) (line AB) shows the disconformity and

samples within overlying (Upper Pleistocene) deposits. Photos: 1: general view with 14m sampled section and position of younger valley fill deposits

(between dashed lines). 2: location of three samples along road cut, the upper of these samples was collected above the disconformity (dashed line) and

shows normal polarity (Agust et al., 1999). Note initial dip toward valley (left) in younger beds. 3: recent excavation (50 m from section) showing

disconformity. 4: details at point g in photo 2, showing unconsolidated conglomerate with subangular intra-basin clasts immediately below discontinuous

calcrete bed (top normal bed of preliminary section). Similar exposures are typical along both sides of the immediate Cullar valley.



5/23

metamorphic grains of southern provenance. North-

directed paleocurrents dominate in this northeastern

margin of the Basin, where small local ponds had

developed. These deposits belong to the same lithostrati-

graphic unit as those in the highest topographic position

across the fault, in the upthrown block. The Puerto Lobo

(PL) paleontological site (Agust, 1984) is located in thetopographically lower (older) part of the down-faulted

block. Higher in the section there are palustrine deposits,

capped by a continuous alluvial unit composed of cross-

bedded gravels and siltstones with paleocurrents towards

the SW. This uppermost unit shows a distinct change in

clast composition and provenance (Mesozoic carbonate

from the north and east) compared to the lower part of the

section (metamorphic grains from the southern moun-

tains).

The site of Huescar-1 (HU-1) (UTM 30543727E,

4183228N, Z 948) is located in the next valley (Barranco

de Las Canadas),o1 km NE of PL (Fig. 2). There are two

other paleontological sites in these younger deposits of the

down-faulted block: Loma Quemada-1 and -2 (LQ-1, LQ-

2), 2 km W (30541848E, 4181276N, Z 900) (Soria et al.,

1987). LQ-1 and LQ-2 have a micromammal fauna similar

to that from PL (Agust, 1984). The site of Huescar-3 (HU-

3) (30543408E, 4182540N, Z 1000) is located close to the

PL section, but is on the other side of the fault in the

upthrown block. The HU-3 fauna are Pliocene in age (Sese,

1989), thus much older than those from sites on the down-

dropped side.

The Puerto Lobo and Barranco de Las Canadas areas

expose a variety of lithologies and show lateral changes in

local facies. Wherever possible, we collected the reddish-brown siltstones for these preliminary magnetostrati-

graphic sections. Good exposures allow for the integration

of biostratigraphic and paleomagnetic data into a litho-

stratigraphic frame. These results offer an opportunity to

identify magnetozone boundaries and to make a distant

correlation to the contemporaneous Cullar section. Pa-

leontological sites with similar age but in different paleo-

environmental and paleogeographical situations allow an

examination of faunal diversity (Alberdi et al., 2001), as

well as a more complete biostratigraphy.

4. Methods

Recent work in the Baza Basin provided a detailed

lithostratigraphic frame for the NE sector, around Orce-

Venta Micena, along with the magnetostratigraphic loca-

tion of the PliocenePleistocene boundary (Scott et al.,

2007). This present study is an expansion both northwards

to Huescar and southwestwards to Cullar. Numerous

stratigraphic sections have been measured in and between

these three areas, permitting some insight into the

complexity of deposition facies and the paleo-topographic

relationships across this part of the Baza Basin. These

different paleo-geographical situations show lateral litho-

logic changes, ranging from the coarse-grained facies

typical of paleo-margins to the fine-grained sulfate-

dominated facies around the central paleo-lake. Although

a detailed lithostratigraphic correlation between these three

marginal sites remains ellusive, the addition of magneto-

zones greatly improves the temporal correspondence.

4.1. Sampling methodology

We chose the most propitious sections in the two areas

to study. These sections have three characteristics: (1)

abundant reddish-brown, fine-grained lithologies for pa-

leomagnetic sampling; (2) ease of correlation to other

nearby sections; and (3) close to or including paleontolo-

gical sites to calibrate the polarity sequence. In 2003, we

collected preliminary samples (n 11) from both sections

(Cullar and Puerto Lobo). This provided an outline of the

polarity zonation and an indication of the magnetic

behavior of the different lithologic types. Additional

samples were collected in 2004 (n 16), expanding the

sections. In 2005, we made more detailed collections

(n 41) extending the sections and refining the position

of each magnetozone boundary.

In the Cullar section, we tested the continuity of the

uppermost magnetozone boundary (R to N) by collecting

samples in stratigraphically equivalent beds along the Oria

road (n 5) (UTM 30559593E, 4158301N, Z 970)

(Fig. 2). We also collected samples (n 4) from a

boulder-sized tilted block, broken from the wall of a

paleo-channel (Fig. 5). The orientation of the remanent

magnetic vector in strata within this rotated block would

be a specific test for the antiquity of remanence. This field

test would indicate which magnetic components wereproduced at an early diagenetic stage and which components

were produced after the bank collapsed into the paleo-

channel. At Puerto Lobo the upper part of the section

showed weakly held remanence and ambiguous polarities, so

we resampled (n 6) equivalent strata in reddish-brown

lithologies (taking advantage of a local facies change) at

Barranco de Las Canadas (just above the site HU-1) (Fig. 2).

4.2. Paleomagnetism

We collected block samples (Cullar n 55, Puerto Lobo

n 13) directly through the sections, in shallow excava-

tions made by hand, and in road cuts. From each oriented

block, at least three specimen cubes (1015 cm3) were cut

and sanded (without water), then cleaned with compressed

air. Measurements were made at the Berkeley Geochronol-

ogy Center with a three-axis cryogenic magnetometer

(noise level 1 1012 Am2) enclosed within a room-sized

magnetostatic shield (average field 350nT). Alternating

field (AF) demagnetization (static three-axis, to at least

12 mT) was followed by thermal demagnetization (in at

least eight steps) starting at 90 1C (non-inductive furnace,

residual fieldo2 nT). Polarity results exclude two unstable

samples at Cullar and three at Puerto Lobo. In general, the

quality of the remanence recorded was excellent for the

ARTICLE IN PRESS



6/23

common light reddish-brown siltstones/claystones (93% of

samples) or was poor for the light gray mudstones,

sandstones and carbonates (7% of samples). All specimens

have a two-component magnetization (Fig. 4) made up of a

low-coercivity/low-temperature component (toward the

modern field direction) and a higher temperature compo-

nent (in the normal or reverse direction). Variable ratios

between the modern and ancient remanence components

are the cause for the variations in demagnetization

response (Fig. 4). The modern field component was

effectively reduced by AF treatment followed by heating

to 200 1C (0001/531, a95 2.81, n 45; goethite only:

000 1/531, a95 2.31, n 44).

A few samples displayed a more complex magnetization,

with higher temperature components showing both normal

and reverse polarity. These samples were adjacent to

magnetozone boundaries, apparently reflecting an integra-

tion of ambient magnetic fields and events. This was most

ARTICLE IN PRESS

Fig. 4. Paleomagnetic specimen data from Cullar section. Orthogonal demagnetization diagrams (vector endpoints during demagnetization) are shown for

specimens from normal (left) and reverse (right) strata. Horizontal projects are shown as triangles, vertical projection as circles, and NRM (4 mT) starting

point as asterisks. Magnetization is in units of A/m. The 200 1C demagnetization step delimits modern (normal) from ancient (normal or reverse)

remanence. Directional changes also shown as stereographic projections (circle is horizontal plane) with solid (open) symbols on lower (upper) hemisphere.



7/23

noticeable in a sample (+28 m) within the middle normal

zone. As a test this block was cut into 17 specimens (as

small as 2 cm3), producing five normal (012/54, a95 8.91),

five reverse (206/56, a95 21.31) and seven intermediate

results. These detailed specimens spanned only 7 cm

vertically; however, the polarities were irregularly distrib-

uted in three dimensions. Since this sample was of typical

lithology (siltstone with immature paleosol textures), weassume that the ambient field changed polarity during the

pedogenic alteration process affecting this bed.

Sample directions were calculated both from furthest

position along great circle trends (shown) and from least-

squared line-fit of vector endpoints. The line-fit results were

similar to the great circle trends (75% have o101

difference), although slightly less consistent and usually

(68%) closer to the modern field. The stratigraphic

sequences of paleomagnetic directions are displayed as

the angular difference (D) between a specimen direction (S)

and the expected normal direction (N) (Hoffman, 1984).

This plot directly reflects the multiple component nature of

these vector data. Calculations of VGP latitudes would

yield the same polarity zonation, but are not shown as an

explicit caution against interpreting minor directional

variations as geomagnetic phenomenon.

4.3. Paleontology

4.3.1. Introduction

In the absence of radiometric dates, rodent biochronol-

ogy is helpful for temporal calibration of a polarity

sequence. Below, we briefly review the history of rodent

biostratigraphy in the Cullar and Puerto Lobo regions and

comment on taxonomic issues. In later sections we discuss

the specimens in detail and then interpret their biostrati-

graphic significance.

For a variety of reasons, some biological and some

probably related to collecting bias, most of the Guadix-

Baza micromammal samples are dominated by arvicolid

rodents. Fortunately, the diversity of arvicolid species in

the Basin, combined with their rapid speciation throughout

the holarctic region during the late Neogene, makes theirremains useful for correlation purposes. Nevertheless, there

is considerable disagreement on fossil arvicolid taxonomy

among European investigators. It is beyond this treatment

to summarize all the pertinent literature. A more compre-

hensive review will be published elsewhere. At this point,

we choose to treat most European Microtus-like species

as members of the genus Microtus, with a common

Mimomys ancestor. Thus Allophaiomys, Stenocranius,

Iberomys, Terricola, etc. are considered subgenera of

Microtus. Also, following van der Meulen (1978), we do

not recognize the subgenus Pitymys in Europe. This name

is reserved for the North American M. (Pitymys) pinetorum

and its relatives, although its ancestor, M. cumberlandensis,

apparently immigrated back into North America from Asia

during the Middle Pleistocene (van der Meulen, 1978).

Fossil relatives may eventually crop up in the Old World.

For many years Arvicola has been considered to represent a

genus separate from Microtus, evolved from the extinct

Mimomys savini. However, a new paradigm of Arvicola

evolution has been proposed (Ruiz-Bustos, 1999) suggest-

ing that Arvicola and Microtus share a common Mimomys

ancestor, with M. savini as a separate sterile Mimomys

lineage. In either case, the progression of evolution in the

first lower molar (m1) of European Arvicola has been

extensively documented (e.g., Heinrich, 1990; Ruiz-Bustos,

ARTICLE IN PRESS

Fig. 5. Field test. Road cut showing channel of conglomerate in the lower Cullar section, with an enclosed tilted block (notebook for scale). Four

paleomagnetic samples from block are shown schematically (sample CU05-31 is of normal polarity). Wall rocks, outside of channel are of the same

lithology (siltstone) as in the tilted block.



8/23

1999; Maul et al., 2000), and the morphology of Arvicola

molars can be valuable for correlation purposes. For

instance, Spanish populations of fossil Arvicola with

relatively small molars and negative enamel differentiation

can be assumed to be older than ones of larger size and

negative differentiation or with positive differentiation

(Ruiz-Bustos, 1999). Likewise, rootless Microtus m1sdemonstrate many features that have both taxonomic

and evolutionary meaning, including complexity of pattern

(number and form of triangles) and enamel differentiation.

However, there are a number of named species in western

Europe with overlapping m1 morphologies, and with

sparse fossil material it is often difficult or impossible to

make a satisfactory species identification. For example,

there are at least five Allophaiomys-type species with m1

morphology that are somewhat advanced over typical

Microtus (Allophaiomys) pliocaenicus from the type

locality of Betfia, Romania (Kormos, 1930; Hir, 1998)

including M. ruffoi (Pasa, 1947), M. burgondiae (Chaline,

1972), M. nutiensis (Chaline, 1972), M. chalinei (Alcalde

et al., 1981), M. vandermeuleni (Agust, 1991), and M.

lavocati (Laplana and Cuenca-Bescos, 2000). We agree

with Ruiz-Bustos (1999) that M. chalinei, originally

described from Cueva Victoria in south-eastern Spain

(and also identified from Atapuerca in north-western

Spain; Cuenca-Bescos et al., 2001), is an Arvicola, though

its dental complexity excludes it from ancestry of modern

A. terrestris or A. sapidus. Microtus vandermeuleni was

recently allocated to the genus Tibericola (Unay et al.,

2001; Agust and Madurell, 2005).

Ruiz-Bustos and Michaux (1976) first reported arvicolid

rodents from the Cullar Baza-1 (CB-1) locality. Agust(1986) proposed a rodent biochronology for the early and

middle Pleistocene, including new material from the

localities near Cullar (CU-A, B, C) assigned to the Upper

Biharian. Localities near Huescar (HU-2, LQ-1 and PL)

were assigned to the Middle Biharian European land

mammal age (ELMA), older than the Cullar Baza and

Cullar localities. Sese (1989) reviewed much of the rodent

material from the Guadix-Baza Basin, and reported

specimens from HU-1. The age suggested for HU-1 was

middle Pleistocene, while PL and LQ were lower

Pleistocene, and CB-1 was Biharian. An attempt to

correlate paleontological localities in the Cullar area with

the geomagnetic polarity time scale (Agust et al., 1999)

placed CB-1 in the Toringian ELMA with CU-A, B and C

beneath, in the Biharian. A normal polarity zone at the

level of CU-C was assumed to be from the earliest Brunhes

(based on geologic observations described in this current

report, site CU-C is out of stratigraphic context and is now

assigned to the late Brunhes). CU-A and -B, in sediments

of reversed polarity, were placed just below the Brunhes in

chron C1r.1r (this current report changes the assignment to

an older chron: C1r.2r). These latter localities were

allocated to the Biharian. An arvicolid rodent biochronol-

ogy for the Guadix-Baza area was proposed (Agust et al.,

1999), with Arvicola cantinaus representing the Toringian

(following Fejfar et al., 1998) and three superposed

Biharian zones, from youngest to oldest Terricola

arvalidens, Allophaiomys burgondiae and Allophaiomys

pliocaenicus. They confirmed the earlier placement of CB-

1 in the Toringian, but did not refer to localities in the

Huescar area. A new biostratigraphic scheme was pro-

posed for the Orce region in a recent comment paperby Agust et al. (2007), but the utilization of paleontolo-

gical sites located in landslide zones (O-2, O-3; Gibert L.

et al., 2006, 2007) and post-erosional deposits (Cu-C)

to build this biostratigraphy makes it of questionable

value.

4.3.2. Methods

We reviewed the existing collection of rodents from the

Cullar and Huescar areas housed at the Paleontological

Institute of Sabadell, along with some specimens provided

by Ruiz-Bustos. Two sets of sites bear on calibration of the

polarity sequence in the Cullar area: Cullar Baza-1, CU-A,

-B, -C. CB-1 produced numerous macro- and microfauna

remains (Ruiz-Bustos, 1976) while Agust (1984) and

Agust et al. (1999) described a few rodents from these

Cullar sites. Four paleontological sites are known from the

Huescar area: HU-1, HU-3, Puerto Lobo and Loma

Quemada (Agust, 1984; Sese, 1989). HU-1 has produced

a large collection of macrofauna with some species of

microvertebrates (Sese, 1989). Puerto Lobo and Loma

Quemada have produced only microvertebrates (Sese,

1989; Agust and Madurell, 2005). HU-3 is located 300 m

SE from HU-1 and is topographically 50 m higher but has

produced a Late Pliocene (Ruscinian) macro- and micro-

fauna (Sese, 1989). This is explained by the presence of anormal fault between these sites (Soria et al., 1987). The

specimens from HU-1 and -3 are not located in the

Institute of Paleontology in Sabadell. For these sites

we will rely entirely on the published faunal account

(Sese, 1989). Other records in the database include both

references from the literature and our independent assess-

ment of material examined in the Institute of Paleontology

in Sabadell.

5. Results

5.1. New paleomagnetic data

Two new polarity sequences are now available in the

Baza Basin: Cullar (81 m) in the southwest and Puerto

Lobo (31 m) in the northeast. The extensively sampled

Cullar section has six major magnetozones: a 1.5 m reverse

zone in the lowest strata, followed by a 7.5 m normal zone,

then a 24 m reverse zone, a short normal zone of 2.5 m,

another long reverse zone (25.5 m), and at the top, a 19 m

normal magnetozone (Figs. 6 and 7). Near the base of the

uppermost normal zone is the paleontological/archaeolo-

gical site CB-1 with abundant mammalian remains and

lithic tools indicating human activity (Fig. 8). Near the

base of the lower long reverse polarity zone would be

ARTICLE IN PRESS



9/23

the projected stratigraphic level of site CU-B (using the

correlation of Agust et al., 1999). A very short normal

zone (o10 cm) was found 3 m above the lowest normal

zone. This cryptozone was inadvertently discovered in the

tilted block used for the conglomerate/tilt stability test.

A short (o50 cm) magnetozone, this time with reverse

polarity, was found within the 2.5 m middle normal

magnetozone.

ARTICLE IN PRESS

Fig. 6. Magnetostratigraphic data for Cullar section. Remanence directions, intensity and low-field susceptibility in relation to stratigraphic position are

shown, with means (open circles) and specimens (solid circles, 3 each). Lithostratigraphic column is shown with position of fossil quarry (bone symbol),

samples (arrows) and samples from ancillary section (C) along Oria road.



10/23

ARTICLE IN PRESS

Fig. 7. Cullar section correlated with GPTS. Geomagnetic polarity time scale (after Gradstein et al., 2004) is shown with correspondence to Cullar

polarity zones, D (calculated from Dec. and Inc.) and lithostratigraphy. Cos (D) sin (Inc N)sin (Inc S)+cos (Inc N)cos (Inc S)cos (Dec N-Dec S)

(N 0001, 571; S specimen direction).



11/23

The more sparsely sampled Puerto Lobo sections have

two magnetozones: a 18 m reverse zone, followed by a 13 m

normal polarity magnetozone (Fig. 9). This lower reverse

zone includes the sites of PL, LQ-1, LQ-2 and HU-1, with

their large collection of micro- and macro-fauna. No

fossiliferous sites have been described from the upper

normal zone; however, we identified lithic artefacts in the

conglomeratic gravel bed that caps this section.

5.2. Magnetochronology

Both studied sections present an almost continuous

sedimentary sequence typical of the marginal facies of an

underfilled basin. No evidence of extensive sedimentary

hiatuses have been identified, with missing time restricted

to levels with immature paleosols and local erosional

surfaces. Some features common to these Baza sections

argue for relatively complete sequences (after McMillan

et al., 2002): high sedimentation rates ($10 cm/kyr) and the

large number of strata. This kind of sedimentary record

facilitates the correlation of the identified polarity zones

with the GPTS (Geomagnetic Polarity Time Scale, after

Gradstein et al., 2004). Using the paleontological data as a

general age guide, as well as other tested polarity sequences

in the Basin (summarized by Scott et al., 2007), these new

polarity zones can be correlated to the Pliocene-Pleistocene

part of the GPTS.

For the Cullar section we correlate the lowest normal

zone to the Olduvai magnetochron (C2n), and the under-

lying reverse zone to the end of the early Matuyama

(C2r.1r). The normal polarity zone in the middle of the

section would then correlate to the Jaramillo subchron

(C1r.1n). The upper (19 m) normal zone, starting just

below site CB-1, would correlate to the Brunhes magne-

tochron (C1n). A very thin normal zone (o10 cm, tilt

tested) identified 3 m above the lowest normal zone could

be the Gilsa subchron (C1r.2r.5n, after Channell et al.,

2002). The other short zone, a reverse within the middle

normal zone is interpreted as C1r.1n.1r (see Guo et al.,

2002). Other possible correlations have significant flaws. If

the long upper normal zone corresponds to the Jaramillo

subchron, then the fauna of CB-1 would be 1.01.1 Ma,

which is probably too old. If the lowest normal zone

corresponds to the Jaramillo subchron (therefore, the

middle normal to the Santa Rosa (C1r.1r.1n) (Singer and

Brown, 2002) or the Kamikatsura subchron (Coe et al.,

2004)), then sediment accumulation would have begun late

in this area, and then proceed very rapidly (425 cm/kyr).

This would require the paleo-Cullar Valley/Fan to change

from a paleo-topographic high to a paleo-topographic low.

ARTICLE IN PRESS

Fig. 8. Lithic artifacts. Flint lithic tools from CB-1 quarry (1987 collection, after Vega Toscano, 1989).



12/23

ARTICLE IN PRESS

Fig. 9. Magnetostratigraphic data, GPTS correlation- Puerto Lobo sections. Magnetic information, lithostratigraphy and correlation to GPTS for the

Puerto Lobo sections have the same conventions as in Figs. 6 and 7.



13/23

This younger model is unsupported by the observed stable

mid-fan deposition environment draining westward into

the nearby paleo-Lake Baza with its distinct, but stable

subenvironments (Gibert et al., 2007).

For the less densely sampled Puerto Lobo sections, we

correlate the upper (13 m) normal magnetozone to the

Brunhes chron (C1n). This is guided by the youthfulness ofthe faunal assemblage from sites HU-1 and PL. We

discount the other possibility of correlating the normal

magnetozone to the Jaramillo subchron, as this would

make these faunas 1.21.3 Ma. This is probably too old, as

the HU-1 and PL micromammals appear distinctly young-

er that those from the Orce-Venta Micena area (Fuente-

nueva-3, Barranco Leon-5) which are of this older age

(Scott et al., 2007).

These new results disagree with some previously

published chronologies for the sites CB-1 or Hu-1, dated

using biochronological methods (Hernandez Fernandez et

al., 2004) and for those dated using the amino-acid

racemization techniques (Torres et al., 1997; Ortz et al.,

2000, 2004).

5.3. Tests for antiquity of remanence

5.3.1. Tilt block test

Different components of magnetization can be acquired

during different events in a samples history. A continuing

challenge in magnetostratigraphic research is to determine

which, if any, of the remanent components were acquired

at the time of deposition, or at least early in the process of

lithification-diagenesis. To examine the antiquity of rema-

nence, the fold and conglomerate tests were developed byGraham (1949). These tests take advantage of unusual

geologic situations where changes have occurred in the

orientation and/or position of newly deposited strata.

Some examples are: rip-up clasts, intra-formational con-

glomerates, bioturbation and slumps. These events physi-

cally rotate the young rock, along with any remanent

direction that was already stably magnetized. If the rock

had not acquired a stable remanence before rotation, it

would now have that opportunity, but in a new orienta-

tion. These types of tests allow a comparison of the

remanent directions (both measured and expected) at the

time of deposition and after the time of change. The shorter

the duration between deposition and rotation, the more

stringent the temporal constraints will be for the acquisi-

tion of remanent components.

In the Cullar section, we found a situation where a

migrating paleo-channel cut into a newly deposited over-

bank deposit, causing blocks to cave off the paleo-channel

wall. These fallen blocks were then rotated and incorpo-

rated into the conglomeratic channel deposit. A boulder-

sized tilted block (60 40 cm) was found within the paleo-

channel, and used for a conglomerate or tilted block test.

The block (Fig. 5) has a rectangular shape and is composed

of reddish-brown sandstone and siltstone (bedding: strike

2851, dip 601N). The blocks original stratigraphic position

can be recognized in the parent sediment located in the

adjacent walls of the paleo-channel. Four samples (six

specimens) were collected from this block, spanning 30 cm

of tilted strata.

Tilted samples show stable magnetic components at

about 901 from the expected paleo-field directions (0001/

+571

or 1801/57

1). These components are also distinct

from those in samples collected from in-situ strata back

from the channel walls. However, after correcting for the

bedding within the block, the sample directions have

reverse inclinations (701, in block coordinates) for three

levels and an approximately antipodal normal inclination

(+601) for the other level (Fig. 11). Using these limited

data, we can identify some additional features, since non-

antipodal paleomagnetic directions can be sensitive in-

dicators to the presence of secondary magnetization (Scott

and Hotes, 1996). Fig. 11 shows that the great circle

through these tilted bi-polar data includes the modern field

direction (within 71). This was unexpected, since reverse

polarity was the ambient field direction both before and for

a long time after the channel-forming event. This means

that among the secondary components, the reverse post-

depositional magnetizations (either pre- or post-tilting)

were not as important as the younger (modern) normal

component. Apparently, the processes of near-surface

alteration (weathering in the modern normal field)

dominate the secondary magnetization in these tilted

samples. A similarly directed, modern secondary compo-

nent is common in the data from most in-situ samples

through out the Cullar section. The magnitude of this

secondary component can be calculated at $20% (relative

to the pre-tilting remanence intensity at 2001) using themethodology in Scott and Hotes (1996). These techniques

also calculate an original average inclination of 651 for

these tilted samples (Fig. 11). Considering the limited

number of samples in this test, this successfully approx-

imates the expected mean inclination of 571.

The observations from this tilted block field-test

produce four conclusions: (1) Much of the natural

remanent magnetization (NRM) was acquired at an early

stage of diagenesis and lithification. (2) No significant

ambient (reverse field) component of remanence was

acquired after the block broke loose and was incorporated

in the paleo-channel deposit. (3) A pre-tilting normal

zone (o10 cm thick) of short duration was found. The

presence of this very thin subzone within the tilted

block is another indication of rapid and early NRM

acquisition. (4) Laboratory demagnetization techniques

(AF followed by thermal) can greatly reduce the secondary

magnetizations in samples, but can still leave a 20%

residual component directed toward the modern (normal)

field.

5.3.2. Stratigraphic continuity tests

Changes in paleomagnetic polarity should be indepen-

dent of lithologic changes, and should be reproducable and

continuous between outcrops. To check the continuity of

ARTICLE IN PRESS



14/23

ARTICLE IN PRESS

Fig. 10. Detailed maps and cross-section profiles. Digital relief maps show the location of fossil sites and cross-sections. Topographic profiles have

geologic cross-sections, with polarity, paleo-current directions, and paleo-environments indicated. The Cullar profile (A-B-C) shows the 6 polarity zones

through 81 m of section. The superseded results ofAgust et al. (1999) are 2 km W and 50 m below the CB-1 quarry. Their normal samples are in post-Baza

Formation deposits from the current episode of erosion. The Puerto Lobo-Barranco de las Canadas section (A-B) shows an upper zone of ambiguous

polarity in PL (diagonal lines), which upon re-sampling 600 m to the north in B. Canadas, was found to be a normal magnetozone.



15/23

the upper magnetozone boundary in the Cullar section, we

collected five additional samples along the Oria road, 1 km

east of the primary Cullar section. Samples were from

stratigraphically equivalent beds; however, fewer palustrine

strata were present owing to a facies change (moving up

the paleo-slope from the ancient Lake Baza). The

change from reverse to normal magnetozone was found

in light brown and reddish-brown siltstones on both sides

of the boundary (Figs. 6 and 12). A less specific test for

continuity of magnetozones across facies involves the

correspondence between part of the Cullar sections middle

(24 m) reverse zone (reddish-brown siltstones) and the

(10 m) reverse zone fragment (Agust et al., 1999) 2 k m

north-west, in more distal, gray mudstones and yellow

sandstones. In our initial Puerto Lobo section, samples

(gray sandstones) above the continuous reverse magneto-

zone had scattered directions (ambiguous polarity). Thus,

we collected six samples from equivalent upper strati-

graphic levels, o1 km north-east in Barranco de Las

Canadas (above the site HU-1). These additional samples

were from reddish-brown siltstones, and indicate a long

(13 m) normal polarity magnetozone (Figs. 9, 10), includ-

ing the uppermost exposures intercalated with cross-

bedded conglomerate.

5.4. Biostratigraphy

Strata from the Cullar area were used in an initial

attempt to calibrate the Early to Middle Pleistocene

boundary by Agust et al. (1999). Our stratigraphic

observations and paleomagnetic results differ significantly

from theirs, and therefore influence the biostratigraphic

evaluations. Paleontologic and biostratigraphic modifica-

tions are noted below. We found one m1 referred to

Terricola arvalidens from CU-C and another m1 of

Castillomys crusafonti from CU-B in the collections of

the Paleontology Institute in Sabadell. Both were figured

by Agust et al. (1999). However, we did not find any

specimens ofStenocranius gregaloides from the Cu-C site as

reported by Agust et al. (2007). As indicated by the

crenulated anteroconid in the drawing (Agust et al., 1999;

Fig. 3.1), the Terricola m1 from Cu-C is from a juvenile

individual. The area of the crenulations was broken off

when we recently viewed it, but the specimen is otherwise

intact. No identifiable molars were found in the collections

from CU-A and none were reported (Agust et al., 1999).

Castillomys crusafontiwas reported from four stratigraphic

horizons, spanning 70 m (Agust, 1986; Agust et al., 1987;

Garces et al., 1997) in the Galera Highway section (15 km

ARTICLE IN PRESS

Fig. 11. Antiquity of remanence field test. Paleomagnetic results from the tilt-block shown in Fig. 5 (4 sampled levels, 6 specimens total). Directions in the

present or in-situ coordinate frame are shown on the left, showing the bedding plane within the block (strike 2851, dip 601N). The nearly antipodal

directions are grouped at 0381, 321 (5 specimens, 3 stratigraphic levels), and at 2261, +541 (1 specimen, 1 stratigraphic level). The ambient (post-tilting)

normal and reverse paleo-field directions are shown as triangles (0001, +571; 1801, 571). Directions in the ancient bedding coordinate frame are on the

right, showing the bedding plane within the block as horizontal (although N is not paleo-north). The concentric circle at 57 1 inclination (+ or ),

represents expected values. Note that the reverse polarity group (1181, 701, a95 8.61, k 100.3) and the normal sample (3411, 571) are not exactly

antipodal. Also, the great-circle connecting these bi-polar data passes close to the modern field direction (diamond), which is the common secondary

direction in other Cullar results). Non-antipodal paleomagnetic directions provide the following parameters: a 631, b 111, g 231, s 0.22 (using

methods in Scott and Hotes, 1996). These calculations indicate 22% of the demagnetized remanence is still held by a secondary component in the direction

of the modern field.



16/23

Fig. 12. Correlation panel for Cullar, Orce and Huescar sectors. Lithostratigraphic and magnetostratigraphic correlations between 3 areas with paleo

(Fuentenueva-1), VM (Venta Micena), BL-5 (Barranco Leon-5), FN-3 (Fuentenueva-3), PL (Puerto Lobo) and HU-1 (Huescar-1). The correlation Cullar t

zone. The correlation Orce to Huescar is based on linking a sequence of 9 lithostratigraphic sections (Gibert L, 2006). The offset on the Puerto Lobo fault

provenance. The correlation Puerto Lobo to Barranco de las Canadas is based on physically following key beds.


17/23

northeast of Cullar). This long-lived species is of little value

for zonal calibration purposes at this stratigraphic level.

As noted above, we found a stratigraphic disconnect

between localities CU-B and CU-C with a 70 m hiatus

between. The site CU-C and the deposits that produced the

purported Terricola arvalidens molar occur in post-

erosional basin sediments (not in the Baza Formation);therefore, the CU-C fossils must be considerably younger

than those from CB-1. This produces a correlational

problem, as M. arvalidens fossils from elsewhere in Europe

occur earlier than Arvicola cantianus and Microtus

brecciensis, the species from CB-1 (Table 1). The anatomy

of the broken arvicolid m1 from CU-C is similar to some

m1 morphotypes from Microtus (Terricola) arvalidens, but

it can also be duplicated in modern pine vole species,

including the extant M. duodecemcostatus from Spain

(Brunet-Lecomte, 1988). Closure of buccal reentrant angle

(BRA) 3 and lingual reentrant angle (LRA) 4, modest

penetration of BRA 4 and LRA 5 forming incipient but

distinct T6-7 in the anteroconid, plus the distinct positive

enamel differentiation, identify the molar piece as from an

advanced species of Microtus. The overall pattern, with

T4-5 confluent certainly suggests the subgenus Terricola.

Brunet-Lecomte (1988; Fig. 16) also showed that a

relatively simple morphology of m1, as demonstrated by

the CU-C specimen, can change with age (wear) into a

more complex molar in modern pine voles. We interpret

the arvicolid specimen from CU-C as a Late Pleistocene or

sub-Recent specimen of M. duodecemcostatus or another

related, morphologically advanced species. Therefore, theTerricola arvalidens zone of Agust et al. (1999) is not

confirmed in the Cullar sections. We do not know what

arvicolids existed directly below the Brunhes-Matuyama

boundary at Cullar, although we can speculate based on

the rodents from Puerto Lobo and Huescar-1.

The arvicolid species from Puerto Lobo and Huescar-1

(Sese, 1989), including Mimomys savini(both localities) and

Microtus huescarensis (HU-1), are consistent with a late

Biharian date, in agreement with Fejfar et al. (1998) and

Agust et al. (1999). Both species were reported by Cuenca-

Bescos et al. (1999) from Trinchera Dolina at Atapuerca, in

sediments of reversed polarity directly beneath the

Brunhes-Matuyama boundary. To date, Arvicola cantianus

( A. mosbachensis of Maul et al., 2000) has not been

recorded from the Matuyama chron. In fact, this species

does not seem to appear anywhere in Europe until about

ARTICLE IN PRESS

Table 1

Faunal list

Compilation made for selected micro-mammals (rodents) and large mammals from Pleistocene paleontological sites in the Baza Basin and in northern

Spain (Atapuerca). Data compiled from Mazo et al. (1985), Alberdi et al. (1989), Garca and Arsuaga (1999), van der Made (1999), Gibert J et al. (2001,

2002), van der Made and Mazo (2001), and Scott et al. (2007).



18/23

0.60Ma (Maul et al., 2000). Although an archaeological

and mammal assemblage with Arvicola cantianus was

initially reported from sediments older than 0.73 Ma at

Iserna la Pineta (Coltorti et al., 1982), new dates place this

association at about 0.61 Ma (Coltorti et al., 2005).

The paleontological sites located in the area of study can

be sequenced from older to younger (Figs. 12 and 13;Tables 1 and 2) as follows: [Early Pleistocene] Cullar

Baza-A and B (CU-A, CU-B); [late Early Pleistocene]

Puerto Lobo (PL), Loma Quemada (LQ) and Huescar-1

(HU-1); [earliest Middle Pleistocene] Cullar Baza-1

(CB-1); and the much younger [Late Pleistocene?] Cullar

Baza-C (CB-C). These sites (except for CU-A, -B) are

located in a higher stratigraphic level than the Lower

Pleistocene paleontological/archeological quarries of the

Orce-Venta Micena sector (VM Venta Micena, BL-

5 Barranco Leon-5 and FN-3) (Table 1). The sites PL,

LQ, and HU-1 are placed between the Jaramillo and

Brunhes polarity zones, indicating a range in age of less

than 200 kyr.

Agust and Madurell (2005) identified Allophaiomys

ruffoi from VM and Allophaiomys lavocati from FN-3,

PL and LQ. Specimens from VM we have viewed and the

m1s illustrated from FN-3 by Agust and Madurell (2005)

can be duplicated in the type material of Allophaiomys

pliocaenicus from Betfia II, Romania, illustrated by Hir

(1998; Fig. 11). The PL and LQ specimens we have

examined are advanced over the simpler m1 morphotypes

reproduced by Hir (1998), and they do not display the

combination of distinctive narrowing of T4-5, tendency

towards closure of the dentine isthmus between T4-5 and

the anteroconid, and the reduced but elongated anteroco-

nid shape that characterizes M. lavocati. For now, until

more specimens have been collected and a quantitative

analysis has been completed, we prefer to view the

specimens from PL and LQ as Microtus sp.The new paleomagnetic and stratigraphic information

presented in this paper, which place the Cullar Baza-1 site

at 0.75 Ma (basically at the Matuyama-Brunhes boundary)

also sets a new date, at least in Spain, for the Toringian-

Biharian ELMA. As noted above, the extinct water vole

Arvicola cantianus has been generally used to define this

boundary (see Fejfar et al., 1998; Maul et al., 2000), and

has not previously been recorded from sediments securely

dated older than 0.61 Ma. But such changes are to be

expected as the stratigraphic and paleomagnetic history of

more basins becomes established. Since LMAs are defined

on the basis of fauna and not, strictly speaking, dates, it is

conceivable that the Toringian-Biharian boundary could

be of different times in different areas of Europe.

Theoretically, within a context of mosaic morphological

evolution, the A. cantianus dental morphology could evolve

at different rates in different areas. However, we think that

once established, a successful arvicolid species would

explosively disperse through Europe within a few hundred

years. The modern introduction and dispersal of the extant

North American muskrat, Ondatra zibethicus, in Europe is

a good example (Danell, 1996). Also, the paucity of

ARTICLE IN PRESS

Fig. 13. Detailed correlation of Matuyama/Brunhes boundaries. The upper Matuyama to lower Brunhes magnetozone boundary correlation for the north

and east sectors of the Baza Basin, shows the diachronous development of sedimentary facies. The palustrine events at Cullar and B. Canadas are not

contemporaneous. These sedimentary events were controlled by paleo-geographic location at Cullar and by neo-tectonic uplift nearby at B. Canadas. Here

climate change was a secondary effect. In the PL section, paleo-currents switch direction between the base and top of the section as a consequence of the

uplift in Botardo hill.



19/23

appropriate sediments for radiometric dating and the

minimal number of long stratigraphic sequences for

magnetostratigraphic determinations in other European

settings can leave considerable flexibility in estimates of the

Toringian-Biharian boundary.

6. Chronostratigraphy and geological implications

Paleomagnetic polarity changes are worldwide events

with the advantage of chronostratigraphic horizons, that of

providing a glimpse into the events and processes happen-

ing at that specific time. Contemporaneous features and

influences such as paleo-geography, paleo-environments,

fauna and neo-tectonics can be compared accurately at

magnetozone boundaries or interpolated between bound-

aries. The Early Pleistocene has boundaries that are

approximated by the limits of two normal polarity chrons,

the Olduvai (C2n) and the Brunhes (C1n). For now, there

are three sections in the Baza Formation with one or more

of these boundary reversals: Cullar, Huescar-Puerto Lobo

(both from this report) and the composite Orce sections

(Scott et al., 2007). These sections represent the processes

of accumulation at three distinct points around the edges of

the Baza Basin, the south-central, the north-eastern and the

south-eastern, respectively, and provide an opportunity for

direct comparison and contrast (Figs. 12 and 13).

As a general observation, the PlioceneLower Pleisto-

cene boundary (top of Olduvai zone) in the Orce area is at

900 asl (Salar Valley; Scott et al., 2007), compared to thesame zone in Cullar, also at 900 m asl. The LowerMiddle

Pleistocene boundary (base of Brunhes zone) in Cullar is at

965 m (asl.) and in Huescar-Puerto Lobo at 955 m (asl.),

but as of yet is not identified around Orce owing to erosion

and/or non-deposition.

Specific comparisons of paleo-environments can begin at

the start of the Pleistocene, when the Orce area, which had

been experiencing a period of protracted lacustrine

deposition, was the site of an advancing fluvial system

coming from an eastern highlands. Meanwhile in Cullar,

the alluvial fan deposition, coming from the southern

highlands continued. As the Middle Pleistocene begins, the

Huescar-Puerto Lobo area had been changing from ponds

and marshes fringing the northern Basin margins to small,

local fans and alluvial drainages coming from an emergent

neo-tectonic structure (Botardo Hill) (Fig. 13). Simulta-

neously at Cullar, palustrine facies was prograding onto

the Cullar fan, with an expansion of palustrine and distal

alluvial environments replacing the mid-fan and alluvial

facies.

Sedimentation rate in fine grain deposits of the Orce area

was $10 cm/kyr (Scott et al., 2007) through the Early

Pleistocene, while in the Cullar area the sedimentation rate

was slower until after the Jaramillo subchron. The Cullar

area experienced an accelerated sedimentation after theJaramillo and into the Brunhes, reaching equivalent rates

to those from the Early Pleistocene deposits of Orce. This

acceleration of sedimentation rate in the Cullar area was a

long-term process interpreted as a product of the prograd-

ing paleo-lake Baza and its marginal valley floor (Figs. 14

and 15). Other processes, such as neo-tectonic uplift and

changing paleo-climates, were secondary variables to the

Cullar sections specific paleo-geographic location. How-

ever neo-tectonic activity, on both the regional and local

scale, was the dominant process at the Huescar marginal

sites. Meanwhile, in the central facies of the paleo-Lake

Baza there was continuous sedimentation in which paleo-

climate was the primary triggering factor (Gibert L. et al.,

2001, 2007; Gibert L, 2006).

The chronology and biostratigraphy of the Baza Basins

Pleistocene fossil vertebrate sites starts in the Orce area,

with MN17 fauna (Fuentenueva-1 quarry) at $1.5Ma,

$25 m above the Olduvai normal zone. The paleontologi-

cal/archeological quarries of Venta Micena (1.3 Ma),

Barranco Leon (1.25 Ma) and Fuentenueva-3 (1.2 Ma)

provide an Early Pleistocene fauna $50 m above the

Olduvai normal zone (Scott et al., 2007). Then the fossil

sites below the Matuyama/Brunhes boundary in the

Huescar area (Puerto Lobo, Loma Quemada and Hues-

car-1) provide a late Early Pleistocene fauna at 0.9Ma.

ARTICLE IN PRESS

Table 2

Chronostratigraphic sequence and biozonation for the Baza Basin

A chronostratigraphic summary is shown for the paleontological sites in

the Baza Basin that can be correlated to the Geomagnetic Polarity Time

Scale. Also shown are the associated continental stages and preliminary

biozones for the Baza Basin.



20/23

ARTICLE IN PRESS

Fig. 14. Age versus sediment accumulation. Sediment accumulation rates are compared between the Cullar and Orce sections, based on magnetozone

boundaries. At the position of the Cullar section, the aggrading valley floor environments (palustrine/lacustrine) arrived soon after the Jaramillo subchron,

which significantly increased the sediment accumulation rate. Projecting this rate to the upper strata, gives an approximate age of 600 Kya. Note that the

top of the Orce sections is an erosion surface, and that the uppermost strata are pre-Jaramillo in age (Scott et al., 2007). Therefore, this figure shows a

minimum accumulation rate for Orce, so that the associated fossil quarries might be slightly older than shown. In Cullar section, the reference

magnetozone boundaries are used as hinge points for accumulation trends in a second refined curve (following McMillan and Stoner, 2005).

Fig. 15. Depositional model for Cullar area. Schematic model which explain changes in sedimentation rate across the Cullar margin during the Late

Pliocene to Middle Pleistocene. Neogene sedimentation starts slowly, with the onlapping alluvial fan at $2.1 Ma. A change to higher sedimentation rates,

owing to the aggrading valley floor, finally reaches the Cullar section about 0.9 Ma as the paleo-Baza Basin continues to fill. This process of net

accumulation stops sometime after 0.6 Ma, and reverses into an erosional mode, that continues today.



21/23

Then immediately above the Matuyama/Brunhes bound-

ary is the paleontological/archeological quarry at Cullar

Baza-1 that provides an initial fauna from the Middle

Pleistocene at 0.75 Ma. And finally, the Cullar Baza-C site

(Agust et al., 1999), which is from post-Baza Formation

deposits, provides a new opportunity to examine fauna

from approximately the Middle/Late Pleistocene boundary($0.1Ma).

7. Conclusions

A study of the lithostratigraphy, magnetostratigraphy

and micro-mammal paleontology was made at two

sections in Quaternary deposits of the Baza Basin.

Magnetostratigraphy of the Cullar section (81 m) has

located both the PliocenePleistocene boundary and the

Lower PleistoceneMiddle Pleistocene boundary. The

paleontological/archeological quarry Cullar Baza-1 is

only 2 m above the Matuyama-Brunhes (LowerMiddle

Pleistocene) boundary. Six major polarity zones have

been identified in this section. The normal zones are

correlated to the Olduvai (C2n), Jaramillo C1r.1n), and

the early part of the Brunhes (C1n) chrons. The base of

Cullar section is dated at $2.0 Ma.

The top of both studied sections are some of the last

deposits before stream piracy captured the drainage of

the previously endorheic Baza Basin. Extrapolation of

sedimentation rates suggests an age for the top of the

Cullar section at around 600 Kya. This indicates that the

opening of the Basin occurred during the MiddlePleistocene, around 600 Kya.

An age of$0.75 Ma can be given to the extensive Cullar

Baza-1 fauna and lithic artifacts.

Magnetostratigraphy of the Puerto Lobo sections

(31 m), 25 km to the north near Huescar has also located

the Lower PleistoceneMiddle Pleistocene boundary.

Two polarity zones have been identified in this section

and are correlated to the latest Matuyama (C1r.1r) and

the initial part of the Brunhes (C1n) chrons.

An age of 0.9 Ma can be given to the paleontological

sites of Huescar-1 and Puerto Lobo.

The change from the Biharian to the Toringian stages

can be approximated by the Matuyama/Brunhes polar-

ity change.

Throughout the Pleistocene, local and regional neo-

tectonic activities have played important roles, along

with the changing climate, in influencing the distribution

of depositional facies and paleo-environments found

within the Baza Basin. However, paleo-geographic

location will determine when, and where these influences

are manifested as sedimentary deposits.

The presence of early humans in the Baza Basin can be

found at 1.31.2 Ma in the Orce area (previous work), at

0.75 Ma in the Cullar area, and again at $0.7 Ma in the

Huescar area.

Acknowledgments

We are grateful to the Paleontological Inst. Crusafont

and Dr. A. Ruiz-Bustos for facilitating the review of the

fossil material. We thank Dr. F. Tortosa for facilitating

detailed digital elevation maps of the area, L. Smeenk for

laboratory assistance and Dr. C. Ferrandez-Canyadell forhis useful comments which improved the manuscript. This

project was funded in part by the generous support of

Earthwatch Institute a National Science Foundation grant

(EAR-0207582) and project CGL2005-05337 BTE from the

Spanish Government.

References

Agust, J., 1984. Bioestratigrafa de los depositos PlioPleistocenos de la

depresion Guadix-Baza (Prov. Granada). Paleontologia i Evolucio 18,

1318.

Agust, J., 1986. Continental mammal units of the PlioPleistocene fromSpain. Mem. Soc. Geol. It. 31, 167173.

Agust, J., 1991. The Allophaiomys complex in southern Europe. Geobios

25, 133144.

Agust, J., Madurell, J., 2005. Los arvicolidos (Muroidea, Rodentia,

Mammalia) del Pleistoceno inferior de Barranco Leon y Fuente Nueva

3 (Orce, Granada), datos preliminaries. In: Toro, I., Agust, J.,

Martnez-Navarro, B. (Eds.), El Pleistoceno inferior del Barranco

Leon y Fuente Nueva 3, Orce (Granada), Monografias de Arqueolo-

gia. Junta de Andaluca, Sevilla, pp. 137145.

Agust, J., Moya-Sola, S., Pons-Moya, J., 1987. La sucesion de mamferos

en el Pleistoceno inferior de Europa: proposicion de una nueva escala

Biostratigrafica. Paleontologia i Evolucio Mem. Esp. 1, 287295.

Agust, J., Oms, O., Pares, J.M., 1999. Calibration of the EarlyMiddle

Pleistocene transition in the continental beds of the Guadix-Baza Basin

(SE Spain). Quaternary Sciences Reviews 18, 14091417.Agust, J., Oms, O., Pares, J.M., 2007. Biostratigraphy, paleomagnetism

and geology of the Orce ravine (Southern Spain). Comment on the

paper by Gibert et al., 2006 [Evaluation of the Olduvai sub-chron in

the Orce ravine (SE Spain). Implications for PlioPleistocene mammal

biostratigraphy and the age of Orce archaeological sites]. Quaternary

Sciences Reviews 26, 568572.

Alberdi, M.T., Alcala, L., Azanza, B., Cerdeno, E., Mazo, A., Morales, J.,

Sese, C., 1989. Consideraciones bioestratigraficas sobre la fauna de

Vertebrados fosiles de la cuenca de Guadix-Baza (Granda, Espana).

In: Alberdi, M.T., Paolo, B. (Eds.), Trabajos sobre el Neogeno-

Cuaternario. Museo Nacional de Ciencias Naturales, Madrid, pp.

347355.

Alberdi, M.T., Alonso, M.A., Azanza, B., Hoyos, M., Morales, J., 2001.

Vertebrate taphonomy in circum-lake environments: three cases in the

Guadix-Baza Basin (Granada, Spain). Palaeogeography, Palaeoclima-tology, Palaeoecology 165, 126.

Alcalde, G., Agust, J., Villalta, J.F., 1981. Un nuevo Allophaiomys

(Arvicolidae, Rodentia, Mammalia) en el Pleistoceno inferior del sur

de Espana. Acta Geologica Hispanica 16, 203205.

Alfaro, P., Delgado, J., Sanz de Galdeano, C., Galindo-Zaldvar, J., Garca-

Tortosa, F.J., Lopez-Garrido, A.C., Lopez-Casado, C., Marn-Lechado,

C., Gil, A., Borque M.J., in press. The Baza fault: a major active

extensional fault in the central Betic Cordillera (south Spain).

International Journal of Earth Sciences, doi:10.1007/500531-007-0213-z.

Brunet-Lecomte, P., 1988. Les campagnols souterrains (Terricola,

Arvicolidae, Rodentia) actuels et fossiles dEurope occidentale.

Ph.D. Thesis, Universite de Bourgogne, Dijon, France.

Calvache, M.L., Viseras, C., 1997. Long-term control mechanisms of

stream piracy processes in southeast Spain. Earth Surface Processes

and Landforms 22, 93105.

ARTICLE IN PRESS

http://dx.doi.org/10.1007/500531-007-0213-zhttp://dx.doi.org/10.1007/500531-007-0213-z


22/23

Chaline, J., 1972. Les rongeurs du Pleistocene moyen et superieur de

France. Cahiers de Paleontologie, 1410.

Channell, J.E.T., Mazaud, A., Sullivan, P., Turner, S., Raymo, M.E.,

2002. Geomagnetic excursions and paleointensites in the Matuyama

Chron at Ocean Drilling Program Sites 983 and 984 (Iceland Basin).

Journal of Geophysical Research 107, doi:10.1029/2001JB000491 .

Coe, R.S., Singer, B.S., Pringle, M.S., Zhao, X., 2004. Matuyama-Brunhes

reversal and Kamikatsura event on Maui: paleomagnetic directions,40Ar/39Ar ages and implications. Earth and Planetary Science Letters

222, 667684.

Coltorti, M., Cremaschi, M., Delitala, M.C., Esu, D., Fornaseri, M.,

McPherron, A., Nicoletti, M., van Otterloo, R., Peretto, C., Sala, B.,

Schmidt, V., Sevink, J., 1982. Reversed magnetic polarity at an early

Lower Paleolithic site in Central Italy. Nature 300, 173176.

Coltorti, M., Feraud, G., Marzoli, A., Peretto, C., Ton-That, T.,

Voinchet, P., Bahain, J.-J., Minelli, A., Thun Hohenstein, U., 2005.

New 40Ar/39Ar stratigraphic and paleoclimatic data on the Isernia La

Pineta Lower Paleolithic site, Molise, Italy. Quaternary International

131, 1122.

Cuenca-Bescos, G., Canudo, J.I., Laplana, C., 2001. La sequence des

rongeurs (Mammalia) des sites du Pleistocene inferieur et moyen

dAtapuerca (Burgos, Espagne). LAnthropologie 105, 115130.

Cuenca-Bescos, G., Laplana, C., Canudo, J.I., 1999. Biochronologicalimplications of the Arvicolidae (Rodentia, Mammalia) from the Lower

Pleistocene hominid-bearing level of Trinchera Dolina 6 (TD 6,

Atapuerca, Spain). Journal of Human Evolution 37, 353373.

Danell, K., 1996. Introduction of aquatic rodents: lessons of the Ondatra

zibethicus invasion. Wildlife Biology 2, 213220.

Fejfar, O., Heinrich, W.-D., Lindsay, E.H., 1998. Updating the neogene

rodent biochronology in Europe. In: van Kolfschoten, Th., Gibbard,

P.L. (Eds.), The Dawn of the Quaternary. Mededelingen Nederlands

Instituut voor Toegepaste Geowetenschappen, Haarlem, pp. 533554.

Garca, N., Arsuaga, J.L., 1999. Carnivores from the Early Pleistocene

hominid-bearing Trinchera Dolina 6 (Sierra de Atapuerca, Spain).

Journal of Human Evolution 37, 415430.

Garces, M., Agust, J., Pares, J.M., 1997. Late Pliocene continental

magnetochronology in the Guadix-Baza Basin (Betic Ranges, Spain).

Earth and Planetary Science Letters 146, 677687.Gibert, J., Gibert, L., Maestro, E., Iglesias, A., 1998. Two oldowan

assemblages in the PlioPleistocene deposits of the Orce region, SE

Spain. Antiquity 72, 1725.

Gibert, J., Gibert, L., Ferrandez-Canyadell, C., Iglesias, A., Gonzalez, F.,

2001. Venta Micena, Barranco Leon-5 and Fuentenueva-3: three

archaeological sites in the early Pleistocene deposits of Orce, south-east

Spain. In: Milliken, S., Cook, J. (Eds.), A Very Remote Period Indeed.

Papers on the Palaeolithic presented to Derek Roe. Oxbow Books,

Oxford, pp. 144152.

Gibert, J., Gibert, L., Vicente, O., Garca, E., Ferrandez, C., Iglesias, A.,

2002. Excavacion de urgencia en el yacimiento paleontologico

Fuentenueva-1 (Orce, Granada). Anuario Arqueologico de Andaluca.

Actividades de Urgencia 1, 135142.

Gibert, J., Gibert, L., Ferrandez, C., Iglesias, A., Gonzalez, F., 2006.

Venta Micena, Barranco Leon-5 and Fuentenueva-3: three archae-ological sites in the Early Pleistocene deposits of Orce, south-east

Spain. In: Ciochon, R.L., Fleagle, J.G. (Eds.), The Human Evolution

Source Book. Prentice-Hall, Upper Saddle River, NJ, pp. 327335.

Gibert, L., 2006. Analisis de facies y magnetostratigrafia de la cuenca de

Baza. Ph.D. Thesis, Univ. Politecnica de Catalunya, Manresa, Spain.

Gibert, L., Ferrandez, C., Scott, G.R., 2001. PlioPleistocene lacustrine

sedimentation in the Baza Basin (SE Spain) and its relation with

climatic shifts. European Space Agency (ESA) 463, 171175.

Gibert, L., Sanz de Galdeano, C., Alfaro, P., Scott, G.R., Lopez Garrido,

A.C., 2005. Seismic induced Pleistocene landslides in deltaic deposits of

the Baza Basin (SE Spain). Sedimentary Geology 179, 279294.

Gibert, L., Scott, G.R., Ferrandez-Canadell, C., 2006. Evaluation of the

Olduvai sub-chron in the Orce ravine (SE Spain). Implications for

PlioPleistocene mammal biostratigraphy and the age of Orce

archaeological sites. Quaternary Science Reviews 25, 507525.

Gibert, L., Scott, G.R., Ferrandez-Canadell, C., 2007. Reply to Comment

by Agust, et al., on the paper [Evaluation of the Olduvai sub-chron in

the Orce ravine (SE Spain). Implications for Plio-Pleistocene mammal

biostratigraphy and the age of Orce archaeological sites]. Quaternary

Science Reviews 26, 573576.

Gibert, L., Ort, F., Rosell, L., 2007. PlioPleistocene lacustrine evaporites

of the Baza Basin (Betic chain, SE Spain). Sedimentary Geology 200,

89116.Graham, J.W., 1949. The stability and significance of magnetism in

sedimentary rocks. Journal of Geophysical Research 54, 131167.

Gradstein, F.M., Ogg, J.G., Smith, A.G., Agterberg, F.P., Bleeker, W.,

Cooper, R.A., Davydov, V., Gibbard, P., Hinnov, L., House, M.R.,

Lourens, L., Luterbacher, H.P., McArthur, J., Melchin, M.J., Robb,

L.J., Shergold, J., Villeneuve, M., Wardlaw, B.R., Ali, J., Brinkhuis,

H., Hilgen, F.J., Hooker, J., Howarth, R.J., Knoll, A.H., Laskar, J.,

Monechi, S., Powell, J., Plumb, K.A., Raffi, I., Rohl, U., Sanfilippo,

A., Schmitz, B., Shackleton, N.J., Shields, G.A., Strauss, H., van Dam,

J., van Kolfschoten, T., Veizer, J., Wilson, D., 2004. A geological time

scale 2004. Columbia University Press, New York.

Guo, B., Zhu, R., Florindo, F., Ding, Z., Sun, J., 2002. A short, reverse

polarity interval within the Jaramillo subchron: evidence from the

Jingbian section, northern Chinese Loess Plateau. Journal of

Geophysical Research 107, doi:10.1029/2001JB000706 .Heinrich, W.-D., 1990. Some aspects of the evolution and biostratigraphy

of Arvicola (Mammalia, Rodentia) in the central European Pleisto-

cene. In: Fejfar, O., Heinrich, W.-D. (Eds.), International Symposium

Evolution, Phylogeny and Biostratigraphy of Arvicolids (Rodentia,

Mammalia). Geological Survey, Prague, pp. 165182.

Hernandez Fernandez, M., Azanza, B., Alvarez Sierra, M.A., 2004.

Iberian PlioPleistocene biochronology: micromammalian evidence

for MNs and ELMAs calibration in southwestern Europe. Journal of

Quaternary Science 19, 605616.

Hir, J., 1998. The Allophaiomys type-material in the Hungarian

collection. In: Martin, R.A., Tesakov, A. (Eds.), The Early Evolution

of Microtus, Vol. 2, Paludicola, pp. 2836.

Hoffman, K.A., 1984. A method for the display and analysis of

transitional paleomagnetic data. Journal of Geophysical Research

89, 62856292.Kormos, T., 1930. Neue wuhlmause aus dem Oberpliozan von Puskpok-

furdo. Neues Jahrbuch fur Mineralogie. Geologie und Paleontologie

Series B 69, 323346.

Laplana, C., Cuenca-Bescos, G., 2000. Una nueva especie de Microtus

(Allophaiomys) (Arvicolidae, Rodentia, Mammalia) en el Pleistoceno

inferior de la Sierra de Atapuerca (Burgos, Espana). Revista Espanola

de Paleontologia 15, 7787.

van der Made, J., 1999. Ungulates from Atapuerca TD6. Journal of

Human Evolution 37, 389413.

van der Made, J., Mazo, A.V., 2001. Spanish Pleistocene Proboscidean

diversity as a function of climate. The World of Elephants-Interna-

tional Congress, Consiglio Nazionale delle Ricerche Rome,

pp. 214218.

Maul, L.C., Rekovets, L., Heinrich, W.-D., Keller, T., Storch, G., 2000.

Arvicola mosbachensis (Schmidtgen 1911) of Mosbach 2: a basicsample for the early evolution of the genus and a reference for

further biostratigraphical studies. Senckenbergiana Lethaea 80,

129147.

Mazo, A.V., Sese, C., Ruiz-Bustos, A., Pena, J.A., 1985. Geologa y

paleontologa de los yacimientos PlioPleistocenos de Huescar

(Depresion de Guadix-Baza, Granada). Estudios Geologicos 41,

467493.

McMillan, D.G., Stoner, J.S., 2005. Optimizing the depth-age transforma-

tion. EOS trans. AGU, 86, Fall Meeting Suppl., Abstract U43A-0826.

McMillan, D.G., Constable, C.G., Parker, R.L., 2002. Limitations on

stratigraphic analyses due to incomplete age control and their

relevance to sedimentary paleomagnetism. Earth and Planetary

Science Letters 201, 509523.

van der Meulen, A.J., 1978. Microtus and Pitymys (Arvicolidae) from

Cumberland Cave, Maryland, with a comparison of some New and

ARTICLE IN PRESS

http://-/?-http://-/?-http://-/?-http://-/?-


23/23

Old World species. Annals Carnegie Museum of Natural History 47,

101145.

Ortz, J.E., Torres, T., Lamas, J.F., Canoira, L., Garca Alonso, L.,

Garca de la Morena, P., Lucini, M., 2000. Datacion de los

yacimientos paleontologicos de la Cuenca de Guadix-Baza (Sector

Cullar-Baza, Granada) y primera estimacion de la edad de apertura de

la cuenca mediante el metodo de racemizacion de aminoacidos.

Geogaceta 28, 109112.Ortz, J.E., Torres, T., Julia, R., Delgado, A., Llamas, F.J., Soler, V.,

Delgado, J., 2004. Numerical dating algorithms of amino acid

racemization ratios from continental ostracodes. Application to the

Guadix-Baza Basin (southern Spain). Quaternary Science Reviews 23,

717730.

Pasa, A., 1947. I mammiferi di alcune antiche breccie veronesi. Memorie

del Museo Civico di Storia Naturale di Verona 1, 1111.

Ruiz-Bustos, A., 1976. Estudio sistematico y ecologico sobre la fauna del

Pleistoceno Medio en las depresiones granadinas. El yacimiento de

Cullar-Baza-1. Trabajos y Monografas del Departamento de Zoologa

1, 1321.

Ruiz-Bustos, A., 1984. El yacimiento paleontologico de Cullar Baza-1.

Investigacion y Ciencia 91, 2028.

Ruiz-Bustos, A., 1999. The role of the water voles (Arvicola, Rodentia) in

the Quaternary. Estudios Geologicos 55, 8995.Ruiz-Bustos, A., Michaux, J., 1976. Le site prehistorique nouveau de

Cullar de Baza-1. Geologie Mediterraneenne 3, 173182.

Sanz de Galdeano, C., Vera, J.A., 1992. Stratigraphic record and

paleogeographical context of the Neogene basins in the Betic

Cordillera, Spain. Basin Research 4, 2136.

Scott, G.R., Hotes, S.A., 1996. The reversal test: an examination of

secondary directions. Geophysical Research Letters 23, 18051808.

Scott, G.R., Gibert, L., Gibert, J., 2007. Magnetostratigraphy for the Orce

region (Baza Basin), SE Spain: New chronologies for Early Pleistocene

faunas and hominid occupation sites. Quaternary Sciences Reviews 26,

415435.

Sese, C., 1989. Micromamferos del Mioceno, Plioceno y Pleistoceno de la

cuenca de Guadix-Baza (Granada). In: Alberdi, M.T., Bonadonna, F.

(Eds.), Geologa y Paleontologa de la Cuenca de Guadix-Baza. Museo

Nacional de Ciencias Naturales, Madrid, pp. 187215.Singer, B., Brown, L., 2002. The Santa Rosa Event: Ar/Ar and

paleomagnetic results from the Valles rhyolite near Jaramillo Creek,

Jemez Mountains, New Mexico. Earth and Planetary Science Letters

197, 5164.

Soria, F.J., Lopez Garrido, A.C., Vera, J.A., 1987. Analisis estratigrafico y

sedimentologico de los depositos neogeno-cuaternarios en el sector de

Orce (depresion de Guadix-Baza). In: Moya-Sola, S., Agust, J.,

Gibert, J., Vera, J.A. (Eds.), Geologa y Paleontologa del Pleistoceno

Inferior de Venta Micena. Institut Paleontologic Dr. M. Crusafont,

Sabadell, pp. 1134.

Torres, T., Llamas, J.F., Canoira, L., Garcia-Alonso, P., Garcia-Cortes,

A., Mansilla, H., 1997. Amino acid chronology of the Lower

Pleistocene deposits of Venta Micena (Orce, Granada, Andalusia,

Spain). Organic Geochemistry 26, 8597.

Unay, E., Emre, O., Erkal, T., Kecer, M., 2001. The rodent fauna from theAdapazari pull-apart basin (NW Anatolia): its bearing on the age of

the North Anatolian fault. Geodinamica Acta 14, 169175.

Vega Toscano, L., 1989. Ocupaciones humanas en el Pleistoceno de la

Depresion de Guadix-Baza: elementos de discusion. In: Alberdi, M.T.,

Bonadonna, F. (Eds.), Geologa y Paleontologa de la Cuenca de

Guadix-Baza. Museo Nacional de Ciencias Naturales, Madrid, pp.

327345.

ARTICLE IN PRESS


gibert et al. (2007) - the early to middle pleistocene boundary in the baza basin (spain)

Documents