Sedimentary and diagenetic markers of the restriction in a marine basin: the Lorca Basin (SE Spain) during the Messinian

ELSEVIER Sedimentary Geology 121 (1998) 23–55

Sedimentary and diagenetic markers of the restriction in a marinebasin: the Lorca Basin (SE Spain) during the Messinian

J.M. Rouchy a,Ł, C. Taberner b, M.-M. Blanc-Valleron a, R. Sprovieri c, M. Russell d,g,C. Pierre e, E. Di Stefano c, J.J. Pueyo f, A. Caruso c, J. Dinares-Turell b,1, E. Gomis-Coll b,

G.A. Wolff g, G. Cespuglio e, P. Ditchfield e, S. Pestrea h, N. Combourieu-Nebout i,C. Santisteban f, J.O. Grimalt d

a CNRS-ESA 7073, Laboratoire de Geologie, Museum National d’Histoire Naturelle, 43, rue Buffon, 75005 Paris, Franceb Institut de Ciencies de la Terra (I.C.T.-C.S.I.C.), Lluis Sole i Sabaris s=n, 08028 Barcelona, Spain

c Department of Geology and Geodesy, University of Palermo, Corso Tukory 131, Palermo, Italyd Department of Environmental Chemistry (C.I.D.-C.S.I.C.), Jordi Girona 18–26, 08034 Barcelona, Spain

e CNRS-LODYC, Universite P. et M. Curie, 4, place Jussieu, 75252 Paris Cedex 05, Francef Departamento de Geoquimica, Petrologıa y Prospeccion Geologica, Universidad de Barcelona, Z.U. de Pedralbes,

08071 Barcelona, Spaing Department of Earth Sciences, University of Liverpool, Bedford Street North, P.O. Box 147, Liverpool L69 3BX, UK

h Geological Institute, Laboratory of Palaeontology, 3, Caransebes, 78344 Bucharest, Romaniai CNRS-ESA 7073, Universite P. et M. Curie, 4, place Jussieu, 75252 Paris Cedex 05, France

Received 16 September 1997; accepted 8 April 1998

Abstract

The Lorca Basin (southeastern Spain) is part of a chain of small marginal Neogene basins located in the structurallyactive Betic area. The Upper Miocene (Messinian) sequence is composed of a thick diatomite-bearing series (TripoliUnit) overlain by the Main Evaporites, analogous to the classical succession that records the main events during theSalinity Crisis in the Mediterranean region. The shallow restricted conditions of this region amplified the sedimentaryresponses to local and global forcings. An integrated approach using sedimentology, micropalaeontology, stable isotopegeochemistry and organic geochemistry has been applied to the Tortonian=Messinian succession of the Lorca Basin, inorder to obtain a continuous record of the environmental changes. The sediments record two major events which affectedthe whole Mediterranean: (1) high levels of productivity that led to the formation of the diatomite-bearing deposits inthe early Messinian (Tripoli); and (2) the Messinian Salinity Crisis with its two major stages, represented by the Haliteand Gypsum Units, both mainly precipitated from marine-derived brines. The rapid reflooding of the Mediterranean bynormal marine waters at the base of the Pliocene did not reach the Lorca Basin, nor other basins of this part of the Beticarea. Instead, continental sediments were deposited as a consequence of the regional uplift of SE Iberia, which startedclose to the Messinian=Pliocene boundary. The most prominent feature of this basin concerns the record of its restrictionby the time of the deposition of the Tripoli Unit, which led to intercalations of precursor evaporitic layers, consisting ofCa-sulphate deposited in sub-aqueous and sabkha conditions, interbedded with diatomites. This alternation of evaporitesand diatomites proves that the Lorca Basin was periodically restricted and reflooded by marine waters, a possible cause

Ł Corresponding author. E-mail: [email protected] Present address: Palaeomagnetic Laboratory, Fort Hoofddijk, University of Utrecht, Budapestlaan 17, 3584 CD Utrecht, Netherlands.

0037-0738/98/$ – see front matter 1998 Elsevier Science B.V. All rights reserved.PII S 0 0 3 7 - 0 7 3 8 ( 9 8 ) 0 0 0 7 1 - 2

24 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

for this being relative sea-level fluctuations in the Mediterranean. This strengthens evidence of diachronism that suggeststhat the onset of the first Messinian evaporitic deposition was not synchronous, but was dependent on bathymetry and localtectonics. High productivity during the early Messinian in this basin is demonstrated by the thick deposits of diatomites.However, stagnation episodes may have occurred during this interval, as suggested by the preservation of high amountsof organic matter (organic-rich shales) and the extent of bacterial sulphate reduction which apparently occurred duringearly diagenesis. The formation of organo-sulphur compounds, replacement of sulphates by carbonates and the high levelsof elemental sulphur are by-products of diagenetic processes occurring in a restricted hypersaline environment. 1998Elsevier Science B.V. All rights reserved.

Keywords: Messinian; Mediterranean; Betic basins; evaporites; organic matter; sulphur; bacterial sulphate reduction

1. Introduction

The restriction that affected the Mediterraneanduring the Messinian (Salinity Crisis) took placethrough a succession of complex hydrologicalchanges which are documented in the sedimentaryrecord by a sequence of marine marls, diatomite-bearing deposits (Tripoli Unit), evaporites and brack-ish deposits (Lago-Mare) (Cita et al., 1978; Hsu etal., 1978; Rouchy, 1982; Rouchy and Saint-Martin,1992). Large basin to basin variations resulted fromthe overprinting of global events (i.e. plate tectonics,glacio-eustatic and climate fluctuations) by regionalorogenic activity (Cita and McKenzie, 1986; Rouchyand Saint-Martin, 1992; Butler et al., 1995). Theshallow peripheral basins were particularly sensitiveto such variations and provide significant informa-tion on chronology and pattern of restriction.

The Lorca Basin is one of the small marginalbasins (200 km2) located in the structurally activeBetic domain (Fig. 1). The Upper Miocene sed-imentary sequence is composed of marine marls,diatomite-bearing deposits (Tripoli Unit) and evapor-ites (Gypsum and Halite Units) (Geel, 1979; Rouchy,1982; Ortı, 1990; Montenat et al., 1990; Benali et al.,1995; Guillen-Mondejar et al., 1995). The relativelyshallow water conditions which prevailed during theMessinian amplified the sedimentary response to lo-cal and=or global environmental changes. The studyof this basin enables a refinement of: (1) the chronol-ogy of the restriction; (2) the sedimentary and dia-genetic processes related to organic-rich and hyper-saline sedimentation; (3) the evolution of depositiontowards continental settings; (4) the role of tectonicson the basin restriction.

Firstly, the bio- and magnetostratigraphy, sedi-

mentology, fossil assemblages, organic matter con-tents and carbonate stable isotopic compositions of areference section (La Serrata), located in the centralpart of the Lorca Basin (Fig. 1A), were studied. Sec-ondly, additional sampling was carried out in severalparts of the section in order to determine the rela-tionships between organic-rich sediments, diageneticcarbonates replacing sulphates and elemental sul-phur. Finally, the information obtained has been con-trasted with observations of several sections locatedin the northeastern sector of the basin (Fig. 1A).

2. Geological background

The Lorca Basin is an intermontane depressionwhich belongs to a system of interconnected Neo-gene basins located in the eastern part of the BeticMountains. These basins are placed within a com-plex fault zone which crosses this orogenic systemover more than 250 km between Almeria and Al-icante (Fig. 1B). Proposed mechanisms of basinformation are: (1) extensional collapse of the oro-genic belt (Platt and Vissers, 1989); or (2) strike-slipmovement along this faulted system, which in turn isinterpreted as a shear zone related to the collision be-tween the African and Iberian plates (Montenat et al.,1987; de Larouziere et al., 1988; Sanz de Galdeano,1990). The Lorca Basin is bounded by two mainfault systems oriented NE–SW (North Betic to thenorth and Alhama de Murcia to the south) and N–S, which remained active during the Neogene andcontrolled the sedimentation (Montenat et al., 1990;Guillen-Mondejar et al., 1995).

The physiography of the basin, which has ex-isted since the Early Miocene, was modified as a

J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55 25

Fig. 1. (A) General geologic map with locations of the studied sections (adapted from Geel, 1979). (B) Regional structural framework(from Montenat et al., 1987).

consequence of a tectonic phase during the lateSerravallian=early Tortonian (Guillen-Mondejar etal., 1995). After this period, sedimentation beganwith early Tortonian shallow water deposits, i.e. con-glomerates, siliciclastics and carbonates, includingcoral reefs. A palaeogeographical differentiation oc-curred during the Tortonian resulting in two distinctzones: a more stable area to the northwest withcontinuous carbonate and coarse-grained siliciclastic

sedimentation, and an elongated depocentre parallelto the southeastern margin.

In the depocentre, which corresponds to the LaSerrata section (Fig. 2), a 1000 m-thick sequence ofmarls accumulated during the Tortonian and earli-est Messinian (Dinares-Turell et al., 1998) (Fig. 2).The overlying Messinian series comprises pre-evap-oritic laminated deposits (Tripoli Unit), about 130 mthick, and the Main Evaporites (Geel, 1979; Rouchy,

26 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55 27

Fig. 3. (A) Panoramic view of the measured sections of the La Serrata, showing the Tripoli Unit with the three lower layers ofsulphur-bearing diagenetic carbonates. The thickness of the interval between layers I and III is 19 m. This view corresponds to the areacomprised between the sites labelled 1a and 1c on Fig. 1A. (B) View towards the reef carbonate complex of the Cejo de los Enamorados,on the SW margin of the basin, showing: (1) the interfingering of the carbonate talus with the Tortonian marls; (2) the progressivethinning of the Tripoli Unit over the talus; and (3) the Tripoli Unit and Main Gypsum Unit onlapping the reef talus deposits.

1982; Ortı, 1990; Ortı and Rosell, 1990; Benali etal., 1995) (Figs. 2 and 3A). The Gypsum Unit (LaSerrata Gypsum), up to 50 m-thick, outcrops moreor less continuously across the basin. Two boreholesdrilled north of the gypsum outcrops crossed a HaliteUnit, 200 m-thick, underlying massive gypsum de-posits, which could be correlated to the Gypsum Unit(IGME, 1982; Ortı, 1990). Thus, deposition of theHalite Unit was interpreted as predating that of theMain Gypsum Unit or to be partly equivalent to thelower part of this unit (Figs. 1 and 2). The area withthe known maximum thickness of evaporites is con-sidered to represent the more subsident part of the

Fig. 2. The La Serrata sections in the central part of the basin showing the lithology, biostratigraphy and magnetostratigraphy(magnetostratigraphy is from Dinares-Turell et al., 1998). The left-hand section is a simplified stratigraphic log of the whole sedimentarysuccession and the right-hand section corresponds to a detailed log of the late Tortonian=Messinian pre-evaporitic succession. Numbers Ito VIII refer to the sulphur-rich carbonate layers.

basin (Ortı, 1990). The thickness of the outcroppingpre-evaporitic formation decreases regularly towardsthe eastern and western margins of the basin. Onlap-ping geometries between the Gypsum Unit and theolder deposits (Fig. 3B) can be deduced from fieldobservations. In the Lorca depression, the evaporitesare covered by continental marls of undifferentiatedlate Messinian to Pliocene age. In the northwest-ern part of the basin, in the area crossed by theGuadalentin River (Fig. 1A), the Tripoli and Gyp-sum Units are not present and instead thick deltaic tofluviatile siliciclastic deposits are found.

The Lorca Basin, as it is for the rest of Neogene

28 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

basins in southeastern Spain, records a cortical upliftsince the Messinian (Docherty and Banda, 1995).This general uplift might have led to the absence ofa record of the Pliocene transgression in this area,after the deposition of the Messinian evaporites.

The Sierra de la Tercia range (Fig. 1A) wassubmerged during most of the Tortonian and earlyMessinian, forming submarine highs covered by largecarbonate platforms, the talus of which locally in-terfingers with the Tortonian=lower Messinian marls(Fig. 3B). These submarine highs were sub-aeriallyexposed during the late Messinian and remained astopographically highs, which separated different ar-eas of evaporite deposition, as shown by the presenceof evaporites in the Lorca depression and on the oppo-site side of the Sierra de la Tercia (Fig. 1A) (Montenat,1973; Rouchy, 1982; Ortı et al., 1993).

3. Sampling and methods

The evolution of the depositional conditionsis constrained chronologically by a biostrati-graphic study (planktonic foraminifera and calcare-ous nannoplankton) together with a palaeomagneticsurvey that was carried out along the reference sec-tion of La Serrata (Dinares-Turell et al., 1998). Bios-tratigraphic data derive from samples taken ca. every5 m up to 900 m and approximately every 1 m inthe upper part of the section (about 300 m), up to thebase of the Main Gypsum Unit.

Estimations of water depth and indications ofbottom water conditions were obtained from quanti-tative data on benthic foraminiferal assemblages byanalyzing the species variation of 300 specimens inthe fraction greater than 125 µm. Quantitative datawere assessed by Q-mode factor analysis, using theCabfac program (Klovan and Imbrie, 1971; Imbrieand Kipp, 1971). Four factors were rotated and theRotated Factor scores are reported in Table 1. Themost significant species indicated by these factorsare: Factor 1 (dominated by Heterolepa praecinc-tata, Spiroplectammina carinata), uppermost slopeenvironment, ca. 150–200 m depth (Parker, 1958;Wright, 1978); Factor 2 (dominated by Bulimina spp.spinose type, Bolivina dilatata), poorly oxygenatedbottom conditions and high supply of organic carbonin the bottom waters (Verhallen, 1991; Sen Gupta

Table 1Varimax factor score matrix

Var. 1 2 3 4

Allomorphina trigona 0.057 0.025 �0.005 �0.016Ammonia beccarii �0.020 �0.113 0.166 0.362Bolivina alata 0.016 0.026 �0.008 �0.001Bolivina dilatata �0.011 0.325 0.105 �0.143Bolivina punctata �0.053 �0.018 0.838 0.048Bulimina spp. (costata type) 0.039 0.064 0.064 �0.047Bulimina spp. (spinose type) �0.007 0.898 0.032 0.000Cancris oblongus 0.016 �0.006 0.002 0.018Cassidulina carinata 0.041 0.003 0.004 �0.007Cassidulina crassa 0.006 0.029 �0.012 �0.006Cassidulina oblonga 0.008 0.024 0.017 �0.004Chilostomella spp. 0.012 �0.001 �0.002 0.004Cibicidoides pachyderma 0.225 �0.025 0.139 �0.129Cibicidoides ungerianus 0.034 0.042 0.051 �0.026Florilus boueanus 0.109 0.100 �0.041 0.817Fursenkoina schreibersiana 0.000 0.016 0.014 0.014Gavelinopsis praegeri 0.010 �0.068 0.310 0.040Globobulimina spp. 0.082 0.020 �0.018 �0.008Gyroidinoides altiformis 0.020 0.002 0.018 �0.012Gyroidinoides neosoldanii 0.112 �0.012 0.026 0.004Gyroidinoides umbonatus 0.085 0.045 �0.013 �0.011Hanzawaia rodhiensis 0.012 0.022 0.011 �0.011Heterolepa praecinctata 0.737 �0.018 �0.198 0.011Hopkinsina bononiensis �0.015 0.059 �0.020 0.019Lenticulina spp. 0.241 0.032 0.002 0.053Marginulina spp. 0.020 �0.001 0.009 �0.001Martinottiella spp. 0.005 �0.004 0.012 0.002Melonis spp. 0.024 0.045 �0.024 �0.003Nonionella turgida 0.020 0.003 �0.009 0.003Oridorsalis umbonatus �0.032 0.042 �0.028 0.280Orthomorphina spp. �0.005 0.005 0.003 0.026Pullenia spp. 0.018 0.040 0.032 0.208Quadrimorphina spp. 0.004 0.005 0.001 �0.001Rectuvigerina spp. 0.012 0.025 �0.003 0.024Sigmoilopsis schlumbergeri 0.016 �0.005 0.006 �0.007Siphonina planoconvexa 0.002 0.005 0.004 �0.006Sphaeroidina bulloides 0.019 0.003 0.000 �0.003Spiroplectammina carinata 0.498 �0.066 0.285 �0.102Stainfothia complanata 0.000 0.009 �0.008 0.038Textularia mexicana 0.127 0.000 0.049 0.031Textularia ponderosa 0.015 0.000 �0.005 �0.001Uvigerina spp. (costata type) 0.159 0.174 0.076 �0.120Valvulineria spp. 0.058 0.042 �0.024 0.061

and Machain-Castillo, 1993); Factor 3 (dominatedby Bolivina punctata and Gavelinopsis praegeri), thelatter species is indicative of an outer shelf environ-ment, ca. 100 m depth (Parker, 1958; Wright, 1978);Factor 4 (dominated by Florilus boueanus, Ammo-nia beccarii), inner shelf environment, ca. 50–60 m

J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55 29

depth (Parker, 1958; Wright, 1978). Diatom assem-blages were studied in 25 samples characteristic ofthe different intervals, with 300 to 500 specimenscounted for each sample according to the methodused by Schrader and Gersonde (1978a).

The palaeomagnetic results are based on a studyincluding 45 sampling sites distributed through theTortonian=lower Messinian marls and the lower thirdof the Tripoli Unit (Dinares-Turell et al., 1998).The data have been correlated with the GeomagneticPolarity Time Scale of Cande and Kent (1995),using the astronomical time scale of Krijgsman et al.(1995) and Hilgen et al. (1995) as age constraints forthe reversal boundaries and biostratigraphic events.

The bulk mineralogy and carbonate contentsof 25 samples from the homogeneous Tortonianmarls, 330 samples from the pre-evaporitic lateTortonian=Messinian deposits, and 30 samples fromthe intra- and post-gypsum marls of the La Serratasection have been analyzed. The carbonate content ofeach sample was measured on 100 mg of powderedsediment using a manocalcimeter (MCM). The bulkmineralogy was determined by X-ray diffraction us-ing a Siemens D-500 instrument (Ni filtered Cu Kα

radiation). Additional samples were obtained fromsections from the northeastern margin (Carretera deCaravaca, Casas del Mellado, Cortijo de las Cole-gialas, Cortijada del Pozuelo) and from a sequenceof organic-rich shales (labelled ORS in Fig. 2) as-sociated with the sulphur-bearing limestone bed III.Clay minerals from some selected samples have alsobeen studied. These are representative of the mainintervals i.e. twelve samples from the homogeneousTortonian=lower Messinian marls, fourteen from theTripoli Unit and eight from the post-evaporitic marls.The clay fraction was analyzed by the method offree-carbonate oriented pastes (fraction less than 2µm in size) which were glycolated and heat-treatedat 570ºC for 90 min (Brindley and Brown, 1980).Fine-grained carbonates, biosiliceous deposits anddiagenetic carbonates were characterized by normallight microscopy and scanning electron microscopy.

Total organic carbon (TOC) and total sulphurcontent were measured on 48 representative samplesusing the analytical procedure described in Rus-sell et al. (1997). Analysis of organic geochemi-cal molecular markers was performed on selectedsamples, utilizing gas chromatography (GC) and

gas chromatography–mass spectrometry (GC–MS)(Russell et al., 1997).

The isotopic composition (18O, 13C) of calcite wasdetermined on 180 samples from the Tripoli Unit andeight samples from marls immediately underlying thisunit. Sampling resolution ranged from 15 to 150 cm,with an average of 50 cm. The isotopic composition ofdolomite was determined on 51 samples coming fromthe opal-CT-rich dolomitic layers, from the TripoliUnit and from the uppermost dolomitic interval un-derlying the Main Gypsum. Five additional samplesfrom the dolomite associated with the precursor gyp-sum interbeds present in the marginal sections wereanalyzed. Stable isotope analyses were performedon both calcite and dolomite. The ‘calcite’ value istaken as corresponding to the fraction of CO2 pro-duced after 20 min reaction at 25ºC with phosphoricacid. ‘Dolomite’ from carbonate mixtures was iso-lated by selective attack with acetic acid 1 N for 20min and later reacted with phosphoric acid at 25ºCfor 4 days. The CO2 gas was analyzed on a triple col-lector mass spectrometer (VG-SIRA 9). The Ž valuesare expressed in ‰ relative to the PDB reference andthe Ž18O values of dolomites are corrected by�0.8‰for the fractionation effect during the phosphoric acidreaction (Sharma and Clayton, 1965).

4. Stratigraphy

The onset of deposition of the marl series underly-ing the Tripoli Unit is not accurately dated, as the firstdatum (Globigerinoides obliquus extremus First Oc-currence, 8.26 Ma) (Sprovieri et al., 1997) is identi-fied only 420 m above the base of the section (Fig. 2).The next biostratigraphical event identified in the se-quence is the First Occurrence (FO) of Globorataliapraehumerosa at 597 m. Between 640 and 775 m,Gt. praehumerosa was not found, but Gld. obliquusextremus and Gt. continuosa are present, indicatingthat this segment represents a repetition of that in the420 and 590 m interval, probably as the result of sub-marine mass redeposition on the platform slope. Thepresence of this repeated interval and the frequentinclusion, in the marls, of large carbonate blocks re-worked from the shelf areas, indicate that these pro-cesses of resedimentation were common during theTortonian (Dinares-Turell et al., 1998) (Fig. 2). Al-

30 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

though the FO of Gt. praehumerosa is known to oc-cur within the chron C4n.2n at 7.69 Ma (Sprovieriet al., 1997), its presence here in a reversed palaeo-magnetic interval (Fig. 2) is interpreted as being dueto the local environmental conditions, which make itsappearance here slightly later than in deeper Mediter-ranean basins. The Tortonian=Messinian boundary(Globorotalia conomiozea and Reticulofenestra ro-taria FO, 7.15 Ma) occurs 25 m below the first typicaldiatomitic level considered as the base of the TripoliUnit and correlates with chron C3Bn (Fig. 2). Theincrease of dextral Neogloboquadrina acostaensis isrecognized 11 m above the base of the Tripoli Unit(Fig. 2). In the upper part of the section, about 20 mbelow the Main Gypsum Unit, the foraminiferal as-semblage is composed only of species reworked fromthe Cretaceous and Eocene.

From these age constraints a first approximationon the variation of sedimentation rates through thestudied series can be proposed. In the interval be-tween the FO of Gl. obliquus extremus and the FOof Gt. praehumerosa, the sedimentation rate can beestimated at ca. 34 cm=ka. In the lower part ofthe Tripoli Unit, the reduced thickness of the chronC3Ar., which is ca. 5 m for an interval of 400 ky(Hilgen et al., 1995), implies a much lower sedi-mentation rate (1.2 cm=ka). A gap in sedimentationoccurring near the base of the Tripoli Unit couldexplain this very low rate although no evident ero-sional surface has been observed in the field. The ageof the base of the Tripoli Unit cannot therefore beaccurately and unambiguously defined.

The absence of palaeomagnetic results (intervalnot sampled) and the lack of biostratigraphical mark-ers in the upper two thirds of the Tripoli Unit donot allow the accurate definition of the age ofthe base of the Gypsum Unit. The lack of indige-nous foraminifera and nannofossils in the continentalpost-evaporitic deposits also precludes the recogni-tion of the Mio–Pliocene boundary.

5. The Tortonian and lower Messinian marls

5.1. Bulk mineralogy

These marls display an homogeneous composi-tion with an average amount of 43% carbonates

(calcite and dolomite), 20% clays, 8% quartz, 1%feldspars and a significant proportion of amor-phous material (organic matter, amorphous silica).Dolomite constitutes about 10% of the bulk carbon-ate. Illite represents around 70% of the clay fraction,the rest being chlorite and smectite in equal amounts.

In the uppermost 50 m, several sandstone bedsand dolostone=opal-CT-rich dolomitic layers, up to50 cm thick, are found intercalated within the ho-mogeneous marls (Fig. 4). The nodular and irregularbedding suggests a diagenetic origin for the opal-CT-rich dolomitic layers. Locally (Carretera de Caravacasection), this interval includes some bioturbated car-bonate mud mounds, up to 1 m in thickness, whichcontain serpulid and bivalve accumulations.

On SEM examination, the opal-CT-rich dolomiticlayers appear mainly without structure, with abun-dant voids due to dissolution of diatom frustules(Fig. 5A); however, scarce well-preserved diatomfrustules are present. Lepispheres of opal-CT arecommon (Fig. 5B). The presence of diatoms and ofcherts that formed from diatom silica remobilization,suggests that several episodes of high biosiliceousproductivity occurred before the deposition of theTripoli Unit. With respect to the underlying marls,the carbonate fraction shows a significant increase inthe percentage of dolomite, which is usually the onlycarbonate present in the opal-CT-rich layers (Fig. 4).

5.2. Organic matter

The TOC values range between 0.06 and 0.6% inthe Tortonian marls, but increase to 1–2% above theTortonian=Messinian boundary (Fig. 6). Total sul-phur contents are also low in the Tortonian marls,and again increase above the Tortonian=Messinianboundary. This increase in sulphur is paralleled byan increase in the abundance of organo-sulphur com-pounds (OSC), though compared to the more or-ganic-rich samples of the overlying Tripoli Unit,these are still of minor importance. There are nomolecular indicators of hypersalinity in this part ofthe section, but almost all the samples contain minoramounts of C25 highly branched isoprenoid (HBI)thiophenes, which indicate that there was an inputfrom certain species of diatoms to these sediments(Volkman et al., 1994).

J.M.R

ouchyet

al./Sedimentary

Geology

121(1998)

23–5531

Fig. 4. Mineralogical composition of the Messinian deposits of the La Serrata section. Note that by comparison with the underlying deposits the mineralogy changes rapidlyin the Tripoli Unit. The percentage of gypsum includes both primary and secondary gypsum resulting from sulphur oxidation. Halite and thenardite are related to surficialprecipitation as efflorescences.

32 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

Fig. 5. Scanning electron microscope (SEM) images of the opal-CT-rich dolomitic layers intercalated within the Tortonian=lowerMessinian deposits. (A) Structureless fabric with elongated ghosts resulting from the dissolution of diatom frustules. (B) Dolomitecrystals cemented by aggregates of opal-CT.

5.3. Micropalaeontological markers

The planktonic foraminiferal assemblage is gener-ally poor throughout the sequence and is essentiallyrepresented by Globigerina bulloides, Neoglobo-quadrina acostaensis (left coiling) and Globigeri-noides spp. Globorotalia menardii (left coiling) issporadically present up to about 900 m. Calcare-ous nannofossil assemblages are poor and affectedby moderate to strong reworking of taxa fromCretaceous–Eocene to Middle Miocene deposits.Reticulofenestrids, Coccolithus pelagicus, Calcidis-cus spp. and Helicosphaera stalis are the most fre-quent autochthonous taxa. Discoaster variabilis, D.surculus and D. brouweri are also present in the bestdiversified assemblages. Ceratholiths and other LateMiocene stratigraphic markers are absent in this inter-val. In the rich and well diversified benthonic assem-blages, only Factor 1 (see definition of factors usedfor environmental interpretation above in Section 3)is dominant in the segment below the repeated inter-val, with small fluctuations. Consequently, the depthof deposition is estimated to be about 150–200 m.The fairly constant sedimentation depth implies thatthe accumulation rate equalled the subsidence rateduring the deposition of sediments in this interval.

A significant change occurs above the repeatedinterval, where species characteristic of Factor 3dominate the assemblages, being indicative of a wa-ter depth of about 100 m, up to just above the FOof Gt. conomiozea. In the upper part of this segment,episodic intercalations of assemblages with high val-

ues of Factor 2 are indicative of periods of poorlyoxygenated bottom water conditions, which appearsimultaneously with the precursor episodes of highbiosiliceous productivity recorded by the opal-CT-rich dolomitic layers.

6. The Tripoli Unit

6.1. Distribution of the deposits

The Tripoli Unit, the lower boundary of which isconventionally placed at the base of the first promi-nent diatomite bed, comprises two main members(Fig. 2).

(1) The Lower Member is composed dominantlyof diatomite beds, several metres thick, interbeddedwith silty marls. The thickness of the unit decreasesfrom 90 m in the basin centre to less than 30 mat the margins, due to the thinning of the marlintervals, whereas the total thickness of the massivediatomites stays constant between 18 and 22 m,except in the most marginal areas to the southwestand northeast where only few diatomite layers, onlya few centimetres thick, occur. These diatomites andmarls are also interbedded with different sediments,e.g. fine- to coarse-grained sandstones, commonlycontaining clasts of basement rocks, fine-grainedlimestones or dolostones, laminated and crystallinegypsum, gypsarenites with fibrous gypsum veins.

The most prominent feature of the Lower Memberof the Tripoli Unit is the presence of intercalated lay-

J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55 33

Fig. 6. Total organic carbon (TOC) and sulphur content of the marls and diatomites in the late Tortonian=Messinian pre-evaporiticinterval. H D levels where biomarker assemblages indicate hypersalinity; D D presence of diatoms indicated by biomarkers (DC insignificant amounts).

34 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

ers of sulphur-bearing deposits, mostly carbonates,and gypsum. The gypsum layers are developed alongthe margins of the basins, especially in the NE areabetween Casas del Mellado and up to about 1 km tothe northeast of Cortijada del Pozuelo, but are notapparent in the La Serrata area, where the gypsumhas usually been replaced by carbonates (Fig. 1). AtCortijada del Pozuelo seven layers of gypsum, fewtens of centimetres to 2 m in thickness, and oneof sulphur-rich carbonate with gypsum relics (fourthlayer from the base) are cyclically intercalated withinthe marls and diatomites (Figs. 7 and 8). The sectionsstudied in this marginal area, especially at Cortijadadel Pozuelo (Fig. 8), record episodic gypsum de-position intercalated within the marine Tripoli Unit.This provides evidence that high salinity depositionoccurred well before the precipitation of the mainMessinian evaporitic unit (Main Gypsum and HaliteUnits). This, and the adjacent Fortuna Basin, are theonly places in the Mediterranean area where suchinterbedding of marine diatomites and gypsum isclearly exposed.

In the central part of the study area, be-tween Casas del Mellado to the northeast and theGuadalentin River to the southwest, the majorityof these gypsum beds grade laterally into sulphur-bearing deposits, mainly composed of carbonates(Figs. 8 and 9). At the La Serrata section, they arelocated below the base of the Gypsum Unit, at 108m (layer I), 100 m (layer II), 88 m (layer III), 87.7m (layer IV), 84 m (layer V1), 56 m (layer VI), 48m (layer VII) and 36 m (layer VIII). These layerscan be correlated between La Serrata and Cortijadadel Pozuelo (Fig. 1), except for the stromatolite-likepeloidal limestone labelled V1 which was not clearlyidentified in the marginal sections. In contrast, layerV2 which is well represented at the margins couldbe represented in the La Serrata section by a thinlayer of powdery carbonate associated with sec-ondary gypsum, located 59 m below the Main Gyp-sum. In the La Serrata area, layer VII correspondsto a fine-grained limestone that overlies a brec-cia composed of sandstone boulders and fragments

Fig. 7. Lithological correlations between the central area (La Serrata section) and sections of the northeastern margin (Casas del Mellado,Cortijada del Pozuelo), illustrating the lateral transition of the main sulphur-bearing diagenetic carbonates to massive gypsum along thebasin margin. The letter g or c after each layer number indicates gypsum or carbonate.

of Tortonian marls. Most of these carbonates weremined for sulphur exploitation in the first half of thiscentury, the most intensely worked being layers IIIand IV. The three uppermost layers are not presentin the central part of the studied outcrops (Car-retera de Caravaca) which corresponds to an areaof greater siliciclastic input, represented by thickbeds of reddish coloured sandstones. These sand-stones sometimes display basal erosional contactsand reduced lateral continuity, and are interpretedas channel-fill deposits intercalated within red marlsand clays. Generally, in the area between Casas delMellado and Cortijada del Pozuelo, the replacementof gypsum by carbonate occurs in different levels.The carbonate layers are more widespread in thelower part of the section (layers I to IV) than in theupper part (VI to VIII). Layers IV and VI, as wellas the stromatolite-like layer V, are usually foundintercalated within slumped diatomite and marl se-quences throughout their areal extension, with layerV itself also being slumped in some areas.

(2) The Upper Member has a maximum thicknessof 33 m. It is composed mainly of silty marlstonesshowing an upward increase in thin layers, severalcentimetres to tens of centimetres in thickness, ofreddish sandstones and containing dispersed thin di-atomitic layers in the lower half of the member. Thetopmost part contains sandstone beds that are thickerthan those of the Lower Member. Below the Gyp-sum Unit, the Upper Member ends with a 2 m-thickinterval, composed of fine-grained carbonate (mostlydolomite) beds (Fig. 3A). This uppermost carbonateinterval is not found in the SW and NE marginal ar-eas where a breccia composed of gypsum, carbonateand sand fragments is locally observed just belowthe Main Gypsum (Cejo de los Enamorados, Cortijode las Colegialas).

6.2. Structure and composition of the sediments

Compared to the underlying marls, the TripoliUnit is characterized by sharp variations in lithol-ogy and mineralogy (Fig. 4). The diatomitic deposits

J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55 35

36 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

Fig. 8. Precursor gypsum layers (arrows) intercalated within the Tripoli Unit (Cortijada del Pozuelo section). The thickness of the intervalbetween the lowermost and uppermost gypsum layers is 65 m.

vary from pure diatomite to diatomitic marlstonesor siltstones (Fig. 10A,C) and are usually laminateddue to variations in the amount of diatoms, silici-clastics and carbonates that commonly correspondto calcareous nannofossils. Some of the laminaeare composed of marly diatomites containing angu-lar flakes of diatomitic laminae, indicating episodesof reworking (Fig. 10B). Layers of dark laminatedchert (composed of massive opal-CT), up to 10cm thick, are commonly intercalated within the di-atomites (Fig. 10A,D).

The composition of the carbonate fraction of theTripoli Unit is variable, dolomite being common andeven the exclusive carbonate component in somesamples. A layer, up to 10 cm in thickness, com-posed of aragonite is present just below carbonatelayer I, whereas thin layers of laminated limestonescomposed either of aragonite or calcareous nanno-plankton are intercalated in the marls and diatomitesbelow carbonate layers II, III, IV and V. In the top-most carbonate beds, the carbonate fraction is com-posed of dolomite, except for the thin and intenselydeformed layers underlying the Gypsum Unit, whichalso contain calcite. Dolomite commonly appears asa cement or as aggregates of small crystals up to 20µm (Fig. 11A,B).

The terrigenous fraction in these sediments ismainly composed of quartz, clays and minor amountsof feldspars. The abundance of smectite is signif-icantly higher than in the Tortonian marls, com-monly reaching 30 to 60% of the clay fraction, whichalso contains illite and minor amounts of Fe-chlo-rites, kaolinite, mixed layers and palygorskite. Small

amounts of pyrite are common in the samples, al-though they are below XRD detection limits.

6.3. Sulphur-bearing carbonates and associatedsediments

These deposits display common sedimentary fea-tures, e.g. high porosity, in-situ brecciation and col-lapse structures which are responsible for markedvariations in thickness, from a few centimetres toabout 1 m.

The carbonate layers are composed of variableamounts of calcite and dolomite, locally associatedwith aragonite. Gypsum was a major depositionalcomponent as indicated by abundant pseudomorphsof gypsum crystals that have been replaced by sparrycalcite and dolomite. Gypsum pseudomorphs are sur-rounded by a dolomicritic matrix and consist of iso-lated or rosette aggregates of lenticular crystals up toa few millimetres in size (Fig. 12) and random or ver-tically oriented aggregates of prismatic or lenticularcrystals up to 4 cm in length. In particular, layer IIIshows large undulations at its bottom and top. Thesemight represent the external shapes of domal aggre-gates of selenite with their basal nucleation cones.Celestite locally constitutes aggregates of centime-tre-sized crystals, whilst barite and chert are presentin layer III. Elemental sulphur appears as dissem-inated granules in the carbonate matrix, millimetreto centimetre nodules or aggregates of large crystalsinfilling cavities after gypsum dissolution and veinsbounding the breccia clasts. Locally, centimetre-sizednodules of alunite are found on top of the marls which

J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55 37

Fig. 9. Sketch summarizing the correlations between the sulphur-bearing deposits and the gypsum throughout the basin.

are in contact with the base of layer III. This mineralis reported in similar settings in the Middle Mioceneof the Gulf of Suez (Rouchy and Pierre, 1987), whereit is thought to have been produced by the reaction ofclay minerals with sulphuric acid derived from the ox-idation of sulphides at a redox boundary. Secondarygypsum may be formed recently by oxidation of el-emental sulphur; the former locally impregnates thesurface of the carbonates at outcrop, or sporadicallyforms discontinuous diagenetic layers.

The carbonate layers are interbedded with cen-timetre- to decimetre-thick diatomite and diatomiticmarl beds, containing thin layers of organic-richshales, in particular below layer III. Millimetre-to centimetre-thick layers of laminated carbonates,pure white in outcrop, are also occasionally inter-

calated within the diatomites associated with thesulphur-bearing carbonates. These laminated carbon-ates are composed either of aragonite (Fig. 13A)or, most commonly, of monospecific and dwarf nan-nofossils (<2 µm in diameter) belonging to thegenus Reticulofenestra (Fig. 13B). Similar accumu-lations of coccoliths have previously been reported inMessinian gypsum deposits of Sicily and Spain (For-tuna Basin) and have been interpreted as resultingfrom algal blooms that occurred under stressful con-ditions of increasing salinity (Rouchy, 1976, 1982).

The gypsum layers representing the marginalequivalents of the diagenetic carbonates consistmainly of nodular to mosaic-type (chicken-wire)structures (Fig. 14). This gypsum results from therehydration of anhydrite that formed by early di-

38 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

Fig. 10. Diatomites. (A) Aspect of the diatomite-rich interval composed of whitish laminated diatomites (d), laminated diatomitic marls(m) and dark laminated cherts (c). (B) Top surface of a diatomitic laminae composed of the accumulation of angular flakes of diatomitefragments. (C) SEM view of a massive diatomite composed of a monospecific assemblage of Thalassionema nitzschioides. (D) SEMview of a chert showing ghosts of carbonate crystals and nannofossils within a structureless matrix.

Fig. 11. Dolomitic sediments. (A) Scanning electron microscope (SEM) view of euhedral crystals of dolomite cementing a diatomite. (B)SEM view of fine-grained dolostones from the uppermost part of the Tripoli Unit, which are composed of small euhedral to subhedralcrystals. Notice the skeletal appearance due to the presence of hollows due to dolomite dissolution.

J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55 39

Fig. 12. Secondary carbonate replacing lenticular gypsum crys-tals (a) disseminated within a fine-grained dolomitic matrix (b).La Serrata, layer III.

agenetic growth within a yellowish fine-graineddolomitic matrix and typifies a formation duringperiods of subaerial exposure (sabkha conditions).Some beds of nodular gypsum are arranged in a se-quence quite similar to those observed in the PersianGulf supratidal settings (Shearman, 1978). This se-quence starts with finely laminated stromatolitic-likedolomicritic carbonates, containing displacive lentic-ular crystals and nodules of gypsum, and ends witha fine-grained dolomitic silt irregularly covering the

Fig. 13. SEM views of the carbonates intercalated within the diatomites. (A) Thin carbonate layer exclusively composed of aragoniteneedles (La Serrata section, base of layer I). (B) Thin carbonate layer (below layer III), exclusively composed of monospecificpopulations of dwarf coccoliths (Reticulofenestra sp.) (Volcan quarry, La Serrata area).

top surface of the gypsum nodules that is, in places,covered by coarse-grained reddish sandstones. Inplaces, elongated nodules form vertically orientatedrows or diverging sheaves which characterize the an-hydrite replacement of selenitic gypsum which grewsub-aqueously, indicating that episodes of sub-aque-ous crystallization occurred during the deposition ofthese calcium-sulphate beds.

6.4. Organic matter

6.4.1. PetrographyMicroscopic characterization of organic matter

(OM) was performed on sixteen samples of marlsand diatomites, representing the different intervalsfrom the pre-Tripoli deposits up to the Upper Mem-ber. The composition of the OM does not varygreatly throughout the section and is typified by thepredominance of coal detritus which forms between48 and 92% of the OM that can be identified petro-graphically. The coal detritus is associated with otherterrestrial debris, i.e. gelified phytoclasts (woodyfragments, non-fluorescent, with no distinguishablecellular structures), wood debris, cuticles, spores,pollen (exclusively coniferous) and, in one sample(218), resin particles. Thus, the proportion of OMdebris of terrestrial origin, as observed under themicroscope, may reach up to 99%. Dinocysts arepresent in low amounts (<10%) or are absent, ex-cept for a diatomite level where they reached 25%,

40 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

Fig. 14. Lithofacies of the precursor gypsum layers character-ized by nodular and chicken-wire structures resulting from theinterstitial displacive growth of anhydrite. Cortijada del Pozuelosection.

although in that case, their morphology suggests thatthey could have been reworked. The proportion ofamorphous organic matter (AOM) is generally low(<12%), except for one diatomite sample whichcontains more than 40% AOM. These palynofaciesdata, showing the large predominance of terrestrialorganic elements, together with the abundance ofreworked macro- and nannofossils, demonstrates thesignificance of repeated continental inputs to basinalsedimentation during the Messinian.

6.4.2. Organic geochemistryA molecular biological marker study was per-

formed on nineteen samples representing differentlithologies through the Tripoli Unit, including marls,sulphur-bearing carbonates, organic-rich shales, di-atomites and silicified diatomites. TOC and S con-tents were also determined for these samples. Adetailed study of biological markers was carried outon selected samples: organic-rich shales, carbonatelayer III, carbonate layer IV and a silicified diatomite22 m above the base of the Tripoli (Fig. 6).

TOCs and total sulphur contents are low (TOC<0.2%; sulphur <0.05%) except for the organic-richshales which lie below carbonate layer III, carbonatelayer III itself and the silicified diatomite (Fig. 6).The organic-rich shales have TOC values between21 and 23% (Permanyer et al., 1994; Benalioulhaj etal., 1994; Benali et al., 1995; Russell et al., 1997)with total sulphur contents of 6 to 7.5% (Russell et

al., 1997), carbonate layer III has a TOC of 0.2% anda total sulphur content of 2.75% whilst the silicifieddiatomite has TOC and total sulphur contents of 0.56and 0.21, respectively (Russell et al., 1997).

The organic-rich shales, carbonate layers III andIV and the silicified diatomite are characterized bythe presence of organo-sulphur compounds (OSC).The silicified diatomite contains abundant sulphur-ized diatom markers (C25 HBI thiophenes), as doescarbonate layer IV. As in the late Tortonian and earlyMessinian marls, this indicates an input to these sed-iments from certain species of diatoms (Volkman etal., 1994). In contrast, the organic-rich shales havea varied and abundant distribution of organo-sul-phur compounds, including C20 isoprenoid alkylth-iophenes and isoprenoid bithiophenes, but no C25

HBI thiophenes. The isoprenoid bithiophenes alsoindicate that the depositional=early diagenetic en-vironment was anoxic and hypersaline (SinningheDamste and de Leeuw, 1987; Russell et al., 1997).These samples contain minor amounts of bis-O-phy-tanyl glycerol ether, which are presumed to derivefrom Archaebacteria and are interpreted as indicat-ing salinities of 50–100 g=l (Teixidor et al., 1993;Russell et al., 1997). The lower part of carbonatelayer III has similar molecular markers to the or-ganic-rich shales but with less abundant organo-sul-phur compounds. These include the C20 isoprenoidbithiophenes, the presence of which implies depo-sition in an anoxic and hypersaline environment. Inaddition to bis-O-phytanyl glycerol ether this samplecontains O-phytanyl-O-sesterterpanyl glycerol ether,which reflects an input from Halococcus spp. and isindicative of salinities greater than 250 g=l (Teixidoret al., 1993; Russell et al., 1997). In contrast, OSCare not detectable in the upper part of the layer.Terrestrial organic markers (in particular long-chainn-alkanes) are present in variable abundances in allof the samples. For the rest of the Tripoli Unit, exceptfor the samples noted above, the preserved organicmarkers are composed mainly of higher molecularweight n-alkanes that derive from terrestrial input tothese sediments.

6.5. Micropalaeontological markers

Planktonic foraminifera are even more rare thanin the underlying deposits and are essentially rep-

J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55 41

resented by Globigerina bulloides, with rare occur-rences of Globigerinoides spp. and Neogloboquad-rina acostaensis. Calcareous nannofossils are presentwith very low specific species diversity, and are essen-tially represented by Reticulofenestrids, Coccolithuspelagicus and Sphenolithus spp. In the benthonicforaminifera assemblage, the dominant Rotated Fac-tors were Factor 4, indicative of shallow water envi-ronment, not in excess of about 50–60 m, and Fac-tor 2, indicative of low oxygen concentrations and=orhigh organic carbon contents in bottom waters.

In the uppermost part of the sequence, the au-tochthonous foraminifera are absent; only reworkedspecies from Cretaceous to Miocene age, espe-cially Globotruncanids (Cretaceous) and Morozow-ellids (Eocene) associated with broken and spatial-ized specimens, are present. Most of the calcareousnannofossils are also reworked from Cretaceous toMiddle Miocene sediments, except for small coc-coliths (4–6 µm in size), namely Reticulofenestraspp. and Coccolithus pelagicus, which occasionallyappear as complete coccospheres. The presence ofthese reworked microfossils implies the erosion ofolder deposits from the Subbetic zone of the BeticMountains which, at present, only outcrop at thenorthern margin of the basin.

The diatom assemblages are mostly composedof marine planktonic species and are dominatedby Thalassionema nitzschioides and Thalassiothrixlongissima. Benthic species are scarce (generallyranging from 0 to 2%). Preservation of frustules isgenerally poor and most are broken, indicating thatmechanical reworking played an important role dur-ing deposition; this is also attested by the brecciationof diatomitic laminae. Hence, estimates of the per-centage abundance and diversity of the assemblagescannot be taken as being representative of the com-position of the autochthonous diatom assemblages.

A total of 53 species were identified in the samples.Based on well preserved forms, three main intervalswere distinguished from the base to the top of thesection (Fig. 15). (1) A lower interval, correspondingto the first diatomite levels. This has a high abun-dance of the marine=brackish species Paralia sulcata(ranging from 1 to 40%), the presence of which in-dicates slight fluctuations of salinity (Hajos, 1986;Fourtanier et al., 1991; Mansour et al., 1995). (2) Themain diatomitic interval, with sporadic increases (up

to 100%) in the abundance of Chaetoceros spp., occa-sionally high amounts (10–50%) of the coastal formActinoptychus undulatus-senarius (Schrader and Ger-sonde, 1978a), and the frequent presence of formsof boreal affinity (Actinocyclus curvatulus, Thalas-siothrix longissima, Rhizosolenia hebetata) (Jouse etal., 1971; Koizumi, 1975; Barron, 1992; Mansour etal., 1995). (3) The Upper Member is characterizedby the presence of freshwater species (Aulacoseiragranulata and Stephanodiscus astrea) (Schrader andGersonde, 1978b; Mansour et al., 1995) together withmarine associations, and the boreal influence is evenmore marked. Although the depositional environmentbecomes more littoral towards the top of the unit, ma-rine inputs persist to the top of the Upper Member.

The first layer of diatomites contains a relativelydiversified assemblage of radiolarians and silicoflag-ellates although species considered as upwellingmarkers are absent (J.P. Caulet, unpubl. data). Thediatomites usually contain fish scales and fragments.It has been suggested that the fish fauna from theLower Member is characteristic of shallow waterconditions and a rather restricted connection with theopen sea (Gaudant, 1989).

6.6. Stable isotopic composition of the carbonates

6.6.1. Calcites (Fig. 16)The Ž18O values of calcites exhibit large varia-

tions between �4.73 and C2.76, with an averagevalue of �2.0‰; this average value is some 3‰lower than the mean Ž18O values from equivalentsediments from Sicily (Pierre et al., 1997) and in theAtlantic (Keigwin et al., 1987). The Ž18O of the car-bonates in Lorca is thus interpreted to reflect marineconditions that have been influenced by dilution byinputs of meteoric water. The Ž13C values are alsohighly variable between �11.33 and C0.29, with anaverage value �1.0 which is considered to representconditions of well-ventilated surface waters.

The very negative Ž18O excursions are interpretedas major dilution events. Four major dilution eventsare observed and correspond to intervals character-ized by siliciclastic interbeds or to samples from theUpper Member. The last dilution event occurs at thetop of the sedimentary succession. The positive Ž18Oexcursions characterized by isotopic shifts, the am-plitudes of which range from 1 to 3‰, are commonly

42 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

Fig. 15. Evolution of the assemblages of diatoms in the Tripoli Unit. The percentage of the dominant species (Thalassionemanitzschioides) and of several significant species is shown.

related to intervals characterized by carbonate lay-ers that arise from calcium-sulphate replacement, inparticular layers I, II, III, IV, V1 and VIII (Fig. 16).These changes of isotopic composition reflect peri-ods of increased salinity which occurred during thedeposition of the Tripoli Unit. The positive Ž18Oshifts occur generally in phase with negative Ž13Cshifts and indicate that, during these events calciteprecipitation occurred from more saline solutionsand that carbon was organically derived. Such rapidchanges in salinity associated with stagnation arealso demonstrated by the intercalation of precursorgypsum layers in the lower Tripoli member.

6.6.2. Dolomites (Fig. 16)The dolomitic fraction is generally mixed with

calcite, except at the top of the section where itforms a 1.5 m-thick layer of pure dolomite. The Ž18Ovalues cover a wide range from �1.63 to C3.72and the Ž13C values show a larger dispersion, from�11.61 to �0.81.

The difference between the Ž18O values of co-existing dolomite and calcite in the mixtures is al-ways positive (0.3 to 5.39). These values are veryclose to the theoretical isotopic enrichment betweencogenetic dolomite and calcite (Fritz and Smith,1970; Matthews and Katz, 1977). The dolomite frac-tion may thus be considered as a very early dia-genetic mineral. The difference between the Ž13Cvalues of coexisting dolomite and calcite is generallyhighly negative, suggesting that the dolomite frac-tion was precipitated from interstitial waters where13C-depleted CO2 was largely provided by the oxi-dation of the organic matter present in the sediments.

The Ž18O values .�1:12 < Ž18O < C3:43/ ofdolomites from the dolostones underlying the maingypsum have been normalized to make them compa-rable to the calcite values by subtracting 3.5‰, whichcorresponds to the average dolomite–calcite oxygenisotopic enrichment during precipitation. These ad-justed dolomite Ž values and the Ž values of calcite,apparently show that the end of the Tripoli Unit is

J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55 43

Fig. 16. Stable isotope composition (18O and 13C) of calcite and dolomite in the upper part of the Tripoli Unit, from carbonate layer IIIto the base of the Gypsum Unit.

marked by a sharp drop in the Ž18O values of both cal-cite and dolomite (Fig. 16) which can be interpretedas being caused by a sharp dilution event which oc-curred before the onset of the evaporitic conditions.

7. The Gypsum Unit

This unit reaches up to 50 m thick and is mainlyformed of laminated, detrital and nodular gypsumwhose petrography has been described previously(Geel, 1979; Rouchy and Pierre, 1979; Rouchy,1982; Ortı, 1990). The major part of the unit iscomposed of clastic gypsum of various grain sizes,associated with laminated gypsum deposits formedsub-aqueously, upon which are superposed variableamounts of nodular gypsum after anhydrite. Numer-ous intercalations of dolostones or dolomitic silt-

stones, millimetre- to decimetre-thick, contain gyp-sum pseudomorphs after displacive halite crystals upto several centimetres in size. A 30 cm-thick horizon,displaying a typical chicken-wire facies, has devel-oped within a whitish magnesitic matrix at about 18m from the base of the unit. This facies is found atthe top of a shallowing-upward sequence which be-gins with laminated gypsum and grades upward intoa flat-pebble conglomerate, before the developmentof the chicken-wire facies. Desiccation cracks arepresent in the lower part of the unit.

In spite of its apparent sedimentological homo-geneity across the basin, the unit often exhibits achaotic structure due to the cumulative effects of com-pressive deformations and of dissolution=collapseprocesses (Fig. 17A). Where the unit is complete andundisturbed, the gypsum is massive or layered. Itsuppermost part becomes enriched in siltstone inter-

44 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

Fig. 17. Collapse features and megabreccias affecting the Gypsum Unit in the La Serrata area. (A) Collapse features leading to chaoticstructure are shown in the right-hand part of the photo. Gypsum quarry. (B) Dissolution surface truncating the beds in the middle partof the Gypsum Unit (Carretera de Caravaca cross-road). The arrows indicate gypsum blocks included in the marls above the erosionalsurface.

calations and grades transitionally into the post-evap-oritic marls. A 5 m-thick interval, characterized bylarge folds, probably due to synsedimentary slides,occurs in the lower part of the unit where former in-tercalations of halite could have formed glide sur-faces (B.C. Schreiber, pers. commun.). However, tec-tonic deformation (chevron folds) is commonly ob-served in the lowermost part. Folds and faults are alsopresent in the top-most part of the unit. The consis-tency of the direction (around 210º) of the fold axiswith the regional dips towards the northwest, arguesin favour of westward tilting of the series during latestMessinian=Early Pliocene.

Abrupt lateral changes occur along the La Ser-rata ridge where well stratified gypsum bodies gradeinto megabreccias composed of gypsum blocks upto 10 m in size, floating in a siltstone and marlstone

matrix (Fig. 17A,B). Locally, the transition from theundisturbed sequence to the highly brecciated zonesinvolves outcrop-scale listric faults with moderateblock displacement, grading to chaotic accumulationsof gypsum blocks (Fig. 17A). Locally (Carretera deCaravaca), the lowermost layer of the Gypsum Unitis truncated down to its base by an erosional surface,which is covered by marls containing large fragmentsof gypsum (Fig. 17B). The brecciated interval is gen-erally overlain by highly deformed gypsum, whichforms the upper part of the unit. Therefore, the brec-cias are interpreted as arising from collapse due toin-situ dissolution=karstification of the Gypsum Unit.The location of the major part of the chaotic brecciasin the lower half of the formation suggests a pos-sible correlation with salt bodies previously interca-lated within this interval which could be equivalent

J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55 45

to the Halite Unit present in subsurface. Dissolutionmay have occurred periodically towards the end ofthe evaporite deposition, as a result of an increase infreshwater influence. The existence of faults cross-cutting the gypsum unit probably favoured the influxof freshwater and dissolution of halite and=or gypsumin the lower part of the unit.

In certain locations (Cortijo de las Colegialas), abed of fluviatile conglomerate, containing boulders ofthe Betic basement and Neogene rocks, is intercalatedin the lower part of the Gypsum Unit. It occurs just ontop of the lowermost metre-thick interval of nodulargypsum, and underlies a chaotic interval. The pres-ence of this conglomerate indicates reworking andcontinental influence during the evaporite deposition.

The silty marls that are interbedded with gypsumlayers and form the matrix of the breccia, containvariable amounts of siliciclastics (quartz, feldsparsand clays) as well as carbonates (both calcite anddolomite). The localized presence of platy dolomiteis noted within the matrix of the breccia. This crys-tal habit of dolomite is usually interpreted as re-sulting from inhibition of crystal growth by someions, mainly Cl�, in highly saline interstitial fluids(Bernouilli and Melieres, 1978). Its formation couldbe related to episodes of hypersaline conditions dur-ing which displacive halite crystals grew.

The TOC content of the Gypsum Unit is very low(0.03%) and the only preserved lipids are a range ofn-alkanes, with a maximum at C31, indicating someterrestrial input to the sediment.

Micropalaeontological examination shows thatthese deposits are devoid of microfossils.

7.1. The post-evaporitic marls

The topmost surface of the Gypsum Unit displayslarge undulations on which the post-evaporitic marlswere deposited. The lowermost part of the formationcontains centimetre to decimetre-thick gypsum lay-ers made of selenites and gypsarenites. This provesthat episodic sub-aqueous precipitation of gypsumoccurred during deposition of the marls. The lowerpart of the formation is composed of sequences(up to 10 cm in thickness) made of sands and,locally, gypsarenites, that grade upwards into car-bonate mudstones. Thicker sandstone beds (some 30cm thick) are intercalated in these sequences.

The average carbonate content of the mudstones(calcite and dolomite) is around 48%, the percentageof dolomite being between 5 and 10% of the bulkcarbonate. The rest of the sediment consists mostlyof siliciclastic components, i.e. quartz, clays andsmall amounts of feldspars. The clay fraction ismore rich in smectite (average value 45%) than theunderlying deposits and also contains palygorskite(average 15%), illite (average 30%) and chlorite.

The carbonate fraction is composed of fragmentsof irregular size with a significant proportion of coc-coliths associated with other nannofossils includingThoracosphaerae, Discoasteridae commonly dis-playing broken arms, Nannoconus, and some largebroken foraminifera. These microfossils are rela-tively abundant and are reworked from older Meso-zoic to Miocene series. Geel (1979) reported thepresence of reworked planktonic forms from theSubbetic Units, but also noted the local occurrence ofin-situ assemblages of Lower Pliocene foraminiferathat we have not found. Our data suggest that thepost-evaporitic deposits formed in a continental set-ting, in which gypsum deposits are thought to be theresult of clastic reworking or processes of precipita-tion in continental brine ponds fed by the dissolutionof the Main Gypsum Unit.

8. Discussion

8.1. Record of the main hydrological changes

The Lorca Basin experienced the same hydro-logical changes that affected the whole Mediter-ranean domain during the Messinian, although thesynchronicity of these events for the different sub-basins remains questionable. The marginal positionof the basin and local tectonics amplified the restric-tion prior to and during the salinity crisis (precursorgypsum layers, diagenesis related to both anoxia andhypersaline conditions, early continentalization).

8.1.1. The evolution of the basin and depositionalenvironments during the Tortonian and earlyMessinian

From the base of the studied section up to the ‘re-peated interval’, the Tortonian marls were depositedin external shelf environments, at a water depth

46 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

at least of 150 m. The frequent slumped intervalsand the presence of repeated stratigraphic intervalsindicate that resedimentation processes were impor-tant in contributing to the deposits in the basin(Fig. 2). A high rate of basin subsidence is indi-cated by the thickness of the marls, that reach upto 1000 m. Above the repeated interval, the waterdepth decreased to about 100 m, with episodic de-velopment of poorly oxygenated bottom conditions.The first episodes of high biosiliceous productivity,expressed by the opal-CT-rich dolomitic layers, ap-pear at about 55 m below the Tortonian=Messinianboundary. A rapid shallowing of the basin to waterdepths of 30–50 m is recorded about 7 m abovethe Tortonian=Messinian boundary. Thus, an impor-tant palaeogeographical change occurred in the lateTortonian and near the Tortonian=Messinian transi-tion, which triggered the restriction of the basin.It seems that an important sedimentary gap, proba-bly of several hundreds of thousand years, occurrednear the base of the Tripoli Unit and is related tothis palaeogeographical change. A change in depo-sitional environment is commonly observed in manyother Betic basins near the transition from Tortonianto early Messinian, generally in relation to tectonicevents (Montenat et al., 1990; Coppier et al., 1990).During this interval, two rapid sea-level drops areexpressed in the nearby carbonate platforms of Cabode Gata, located east of Almeria, by erosional sur-faces and megabreccias (Franseen et al., 1998). Inthe Lorca Basin, further studies are needed to con-strain the age, significance and causes of this event,in particular the stratigraphic gap. This change mayhave been predominantly controlled by tectonic ac-tivity which is well-documented at a regional scale,although a global eustatic drop recorded in the At-lantic Ocean (Zhang and Scott, 1996) could haveamplified the effects of tectonics and contributed tothe generation of the sedimentary gap.

8.1.2. Depositional conditions of thediatomite-bearing sequence (Tripoli Unit)

The deposition of these sediments took placein shallow-water environments that underwent rapidchanges of water depth and salinity. These variationsare supported by the intercalation of hypersalineevents as evidenced by the presence of: (1) gypsumlayers preceding the main gypsum precipitation; (2)

organic markers for hypersalinity; and (3) the sharpvariations in the stable isotopic composition of thecarbonates indicating fluctuating salinities.

The significant development of biosiliceous de-posits implies high organic productivity during theMessinian in the Lorca Basin. The age of the baseof the Tripoli Unit cannot be evaluated accurately, asthe base could occur in the time interval correspond-ing to a sedimentary gap, but precursor phases ofenhanced biosiliceous productivity, as shown by theunderlying opal-CT-rich dolomitic layers, occurredas early as in the uppermost Tortonian. Except for theChelif Basin in Algeria where the Tripoli may reachup to 200 m in thickness (Perrodon, 1957; Rouchy,1982), the Lorca Basin contains the thickest TripoliUnit of the Mediterranean basins. The increase ofnutrients required to induce high productivity is usu-ally ascribed in the rest of the Mediterranean basinsas being due to either: (1) upwelling of deep waters(McKenzie et al., 1979; Rouchy, 1982, 1986; Ger-sonde and Schrader, 1984; Muller and Hsu, 1987);(2) input of Atlantic intermediate waters (Bensonet al., 1991); or (3) increased supply of continentalwaters (Van der Zwaan, 1979; Howell et al., 1988).

The episodes of enhanced productivity occurredin shallow and unstable conditions within the LorcaBasin, so that it is unlikely that they were causedby upwelling of deep nutrient-rich waters. More-over, the water depths over the silled areas whichseparated the Lorca depression from the rest of theMediterranean basin were too shallow to permit theentry of intermediate Mediterranean waters. Thus,there are two hypotheses for explaining these condi-tions of enhanced productivity.

The first hypothesis is that nutrient-rich waters,generated outside of the Lorca Basin in deeper areasof the Mediterranean, by upwelling or by influenceof intermediate oceanic waters, were periodicallybrought into the marginal Lorca depression by lat-eral advection (Fig. 18). Studies of the present-dayupwelling system of the Namibian coast show thatthe area of diatomite sedimentation does not corre-spond exactly to the area of deep water resurgence,but it occurs distally on the shelf (Calvert and Price,1983). Geological reconstructions cannot show theexact timing and location of water mass transfer dur-ing the Messinian in the Mediterranean realm, butit is possible that the inputs of marine waters were

J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55 47

Fig. 18. Scenario for the alternating sedimentation of diatomites and precursor gypsum layers, together with related diagenetic processes.

balanced by the high rate of evaporation, as shownby the periodical evolution towards hypersaline con-ditions.

The second hypothesis could explain high pro-ductivity in terms of a contribution of terrestrialnutrients. This concurs with the presence of sig-nificant amounts of terrestrial organic debris andreworked microfossils from older deposits. Further-more, the Ž13Ckerogen of the organic matter showsa shift towards more negative values from belowthe first appearance of Tripoli-type deposits (M.Russell, A. Harriman and G.A. Wolff, unpubl. re-sults). This might reflect a relative increase in ter-restrial supply of dissolved silica and other micro-or nanonutrients, which in turn could have triggeredthe high biosiliceous productivity. Freshwater dis-

charges from the deltaic systems of the NW part ofthe basin could have brought large amounts of ter-restrial nutrients to the whole basin. This hypothesisis strengthened by the fact that the source area ofmost of the reworked microfossils observed withinthe Tripoli Unit, the Subbetic Basement Units, cor-responds to this north to northwestern margin of thebasin. However, it is only at 65 m below the GypsumUnit that the main organic lipid markers indicate ahigher contribution of terrestrial material. This couldbe due to mixing with marine-sourced organic matterin the lower samples.

8.1.3. Precursor evaporitic episodesBiostratigraphical data and the presence of gyp-

sum layers intercalated with the diatomites on the

48 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

margins of the basin, show that the evaporitic con-ditions first started in the Lorca Basin some timebefore the onset of the Main Evaporite depositionin the main Mediterranean basin. The gypsum bed(layer I) lies only 5 m above the N. acostaensis dxdatum, estimated at 6.46 Ma; thus the first recordof evaporitic conditions can be dated at about 6.35Ma, approximately 0.3 million years before the baseof the Calcare di Base in the Falconara section ofSicily (Sprovieri et al., 1996). It is of interest tonote that precursor evaporitic beds are also reportedin some marginal sections of the Central SicilianBasin around 6.3 Ma (Caruso et al., 1997). Similarinterbedded diatomites and gypsum layers have beenreported in the Fortuna Basin, located northeast ofLorca, though accurate correlation has yet to be con-strained (Santisteban, 1981; Muller and Hsu, 1987;Ortı et al., 1993). In the rest of the Messinian basins,precursor calcium-sulphate deposits, sometimes re-placed by carbonates, were reported in the topmostpart of the Tripoli, as in the Sicilian Calcare di Base(Ogniben, 1957; Schreiber, 1974; McKenzie et al.,1979; Rouchy, 1982; McKenzie, 1985; Bellanca andNeri, 1986; Decima et al., 1988; Butler et al., 1995).They appear at ca. 6.05 Ma in the central parts of theSicilian Basin (Sprovieri et al., 1996), but earlier inits marginal areas (Butler et al., 1995; Caruso et al.,1997); marine diatomites were observed interbeddedin the lower part of the gypsum units in Cyprusand Crete (Rouchy, 1982). These data confirm thatthe onset of Messinian evaporitic deposition in thedifferent basins was not simultaneous throughout theMediterranean.

The sub-aerial conditions that affected themarginal areas during the deposition of these cal-cium-sulphate layers (sabkha conditions) impliesthat the base level dropped repeatedly, probably inresponse to the isolation of the basin (Fig. 18). Thealternation of marine flooding and restriction of thebasin, which is required for the deposition of thediatomites and evaporite precipitation, may only beexplained by relative sea-level fluctuations of theMediterranean reservoir, although the hypothesis ofclimatic fluctuations cannot be totally precluded.The time span between the events is too short forthese palaeogeographical changes to be controlledby subsidence or by tectonic processes (Fig. 18). Theshallow and restricted conditions in the Lorca Basin

amplified the eustatic signal, which is not recordedin the deeper basins where more stable hydrologicalconditions prevailed during this interval.

8.1.4. Transition to perennial evaporitic settingsAs previously reported by Benalioulhaj et al.

(1994) and Benali et al. (1995), freshwater inputsincreased during the deposition of the Upper Mem-ber of the Tripoli Unit. This is shown by the dis-appearance of the indigenous foraminifera and theabundance of reworked forms. The stable isotopevalues of the carbonates and the presence of ma-rine diatoms or monospecific and dwarf nannofossilsargue for persistent, but reduced inputs of marinewaters. At the end of the deposition of the unit(uppermost carbonate layers), a sharp dilution eventoccurred, as indicated by the shifts towards low Ž18Ovalues of the carbonates.

As seen above, the Halite Unit, in subsurface,is interpreted as underlying the Main Gypsum Unit(Fig. 2) or to be intercalated in the lowermost part ofthis unit. The latter explanation is supported by theintensely deformed contact of the gypsum with theunderlying deposits which could have been inducedby the dissolution of halite (Ortı, 1990).

The composition of the fluid inclusions in thehalite and the isotopic composition of the sulphatesin the lower part of the Halite Unit of the LorcaBasin indicates that the lower part of the Halite Unitprecipitated from brines of marine origin (Ayora etal., 1994; Garcıa-Veigas et al., 1994, 1995; Cendonet al., 1997).

The dissolution=reprecipitation of previously de-posited halites by meteoric waters is evidenced in theupper part of the Halite Unit (Garcıa-Veigas et al.,1990, 1994, 1995; Ayora et al., 1994). The fluviatileconglomerate observed in the marginal successionof Cortijo de las Colegialas could be related to acontinental episode which occurred during this inter-val of time. The upper part of the Halite Unit hasbeen interpreted (Ayora et al., 1994) as recording thecomplete closure of the basin to any marine waterinput.

After this first evaporitic event, marine waters re-entered the basin and the Gypsum Unit precipitatedfrom marine brines, as shown by the stable isotopiccomposition of the sulphates (Rouchy and Pierre,1979; Pierre, 1982; Garcıa-Veigas et al., 1990;

J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55 49

Utrilla et al., 1992). Moreover, the Gypsum Unit ex-tended laterally beyond the limits of the underlyingdeposits (Halite Unit, Tripoli Unit) locally onlappingthe talus of the Tortonian=lower Messinian carbon-ate platforms (Fig. 3B). This transgressive pattern ofthe gypsum deposits is common in many Messinianbasins where former depressions were filled by haliteor by massive gypsum accumulations (Decima andWezel, 1971; Rouchy, 1982). Thus, after filling ofthe depressions by early deposits, the flattened basinbottom was covered by thin brine layers which pro-gressively extended farther onto the margins duringdeposition of the Gypsum Unit. This sedimentationpattern implies low clastic input to the area and thatthe extent of sedimentation mostly matched that ofsubsidence. The gypsum precipitation mainly tookplace sub-aqueously in shallow ponds submitted toepisodic desiccation, although nodular anhydrite re-lated to early diagenetic growth and intraformationaldissolution suggests that the basin dried up episodi-cally during this evaporitic stage.

8.2. Final transition to continental settings

The influx of marine waters finally ceased atthe end of the gypsum deposition, resulting in theestablishment of freshwater environments. Severalother basins of the Betic domain (Mula, Totana, For-tuna) had evolved earlier to continental conditionsand were not reflooded by marine waters (Monte-nat, 1977; Santisteban, 1981; Rouchy, 1982). Forexample, in the Fortuna Basin, the transition to acontinental environment occurred in the uppermostcycles which form the Main Gypsum Unit in thisbasin (Santisteban, 1981; Ortı et al., 1993). Thebasins of this part of the Betics are different fromother peri-Mediterranean basins (Sicily, Algeria, Io-nian Islands, Cyprus) where open marine conditionswere abruptly restored during the earliest Pliocenetime. This may be explained by the regional uplift ofthis area, near the Mio–Pliocene boundary (Dochertyand Banda, 1995).

The transition towards permanent lacustrine set-tings caused collapse by dissolution of the gypsumand=or halite bodies which underlie the GypsumUnit, or are intercalated within its lower part, re-sulting in collapse breccia textures (Fig. 19). Theintense tectonic deformation in some parts of the

Gypsum Unit suggests that the percolation of fresh-water was enhanced by tectonic fractures. Simi-lar episodes of erosion=dissolution truncating theMessinian evaporites at the end of the main evapor-itic event have been reported in many other basins,such as the Vera Basin, offshore Spain (Fortuin et al.,1995), Ionian Islands, Crete and Cyprus (Rouchy,1982; Rouchy and Saint-Martin, 1992; Delrieu etal., 1993), Corsica Channel (Aleria Group, 1980),and Israel (Druckman et al., 1995). An importanterosional surface has also been reported in the Cabode Gata carbonate platforms, truncating the upperpart of the Terminal Complex which is thought tobe formed simultaneously with the Upper Evapor-ites of the deep Messinian basins (Franseen et al.,1998). All of these support that a major erosionalevent, the effects of which were probably amplifiedin the Lorca Basin by regional uplift, occurred in theMediterranean at the transition between the SalinityCrisis-related deposits and the late Messinian post-evaporitic continental sediments.

8.3. Sedimentary and diagenetic record of thestagnation

During the episodes of high biosiliceous produc-tivity large quantities of organic matter were de-posited, with a significant contribution of diatoms.Thin layers of organic-rich shales formed just beforethe deposition of certain sulphur-bearing carbonates,especially layer III, and their lateral gypsum equiva-lents. These organic-rich shales may represent eitherconditions of higher productivity, or more efficientpreservation. The distribution of organic compoundsof the organic-rich deposits indicate deposition in hy-persaline conditions, which developed before evap-orite precipitation (overlying carbonate or gypsumlayers). The diatom assemblages of the interbed-ded diatomites do not display significant variationscompared to the marine associations of the otherintervals. Blooms of dwarf and monospecific coccol-ithophorids (Reticulofenestra) are thought to char-acterize stressed conditions related to hypersalinity,as in similar assemblages previously reported ingypsum deposits of Fortuna (SE Spain) and Sicily(Rouchy, 1976, 1982).

Poorly oxygenated to anoxic conditions devel-oped during periods of hypersalinity. These favoured

50 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

Fig. 19. Scheme summarizing the main depositional events recognized by this study in the Lorca Basin.

intense diagenesis due to bacterial sulphate reduc-tion in organic-rich sediments, as shown by thereplacement of gypsum by carbonates, the low Ž13Cvalues of the replacing carbonates, and the abundantelemental sulphur included in the carbonates. Theorganic-rich shales are also characterized by signifi-cant amounts of organo-sulphur compounds resultingfrom the early sulphurization of organic matter dur-ing bacterial sulphate reduction. The latter reaction isthought to be responsible for the carbonate replace-ment of gypsum as well as for large modifications

of the original structure of the layers (secondaryporosity, brecciation, collapse, etc.). Later diageneticprocesses led to the formation of nodules of alu-nite. Similar transformations have been described inother geological settings characterized by the pres-ence of calcium-sulphates and organic-rich deposits,such as in the Messinian of the Central SicilianBasin and the Gulf of Suez (Dessau et al., 1962;Pierre, 1982; Rouchy, 1982; Rouchy and Pierre,1987; Pierre and Rouchy, 1988). Sulphate replace-ment by carbonates and other evidence for bacterial

J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55 51

sulphate reduction are characteristic of the centralarea of the Lorca Basin where sub-aqueous condi-tions persisted and a significant amount of organicmatter was accumulated and preserved during earlydiagenesis. In contrast, the intensity of these pro-cesses decreased progressively towards the marginsof the basin, where evaporites formed in shallowerwater settings that were affected by episodes of sub-aerial exposure corresponding to sea-level lowstands(Fig. 18).

9. Conclusions

The integrated approach, incorporating sedimen-tology, micropalaeontology, stable isotope geochem-istry and organic geochemistry, has allowed us tocharacterize the main environmental changes that oc-curred in the Lorca Basin during the Messinian re-striction. This basin experienced the main events thataffected the whole Mediterranean during this period,i.e. the increase of productivity during the pre-evap-oritic interval and the Salinity Crisis, represented bytwo evaporitic units, a Halite Unit and a GypsumUnit. The sedimentary expressions of these events aresuperimposed on progressive shallowing and restric-tion events, which were predominantly controlled byregional tectonics. Because of the shallow restrictedconditions that prevailed in the Lorca Basin, the pre-evaporitic series also recorded the effects of relativefluctuations of Mediterranean sea-level, unlike thedeeper basins, such as those of Central Sicily wherehydrological conditions were more stable.

One of the most significant features observedin the Lorca Basin, is the presence of precursorepisodes of gypsum precipitation intercalated withinthe diatomite-bearing deposits of the Tripoli Unit,that pre-dated the deposition of the main evaporiticunits. Only variations of the Mediterranean sea level,probably controlled by eustatic sea-level fluctuations,can explain such repeated phases of closure andreflooding of the Lorca Basin. This strengthens thehypothesis that the sedimentary record of the onsetof the evaporitic deposition occurred diachronouslyduring the Messinian in the different basins, due tolocal bathymetric and tectonic controls.

The depositional conditions were characterizedby high organic matter fluxes to the sediments due

to high productivity, and by highly concentratedbrines. Such conditions generated, in the centralareas of the basin, anoxia and intense bacterial sul-phate reduction, which were responsible for early di-agenetic sulphurization of functionalized lipids withreduced inorganic sulphur. The process of sulphur-ization leads to enhanced preservation of organicmatter and thus gives information on organic inputand depositional environment (Sinninghe Damste etal., 1989). A multistaged diagenesis that modifiedthe primary mineralogy of the deposits (carbonatereplacing calcium-sulphates and formation of ele-mental sulphur as a by-product) occurred in therelatively deeper areas of the basin, where organicmatter was preserved.

The main evaporitic deposits, i.e. the Halite Unitand the overlying Gypsum Unit, formed from marinebrines. The halite deposition, however, was termi-nated by a significant episode of freshwater dilutionbefore the restoration of dominantly marine condi-tions and deposition of the Gypsum Unit. Similarfreshwater dilution events, which occurred at the endof the first phase of evaporitic basin infilling, are alsoreported in Sicily and Cyprus between the Lowerand Upper Evaporites (Decima and Wezel, 1971;Rouchy, 1982).

The major difference between some Betic basins,including the Lorca Basin, and most other Mediter-ranean basins (e.g. North Africa, Sicily, Cyprus)concerns the post-evaporitic sedimentation, whichwas characterized by permanent freshwater environ-ments in Lorca, whereas open marine conditionswere abruptly restored elsewhere in the Mediter-ranean in the earliest Pliocene. This difference re-sulted from a regional uplift in this part of the Beticdomain, that led to significant differences in therecord of the Messinian=Pliocene boundary in thevarious basins of the Betic domain. The establish-ment of freshwater environments, immediately afterthe evaporitic deposition, was responsible for theoccurrence of a significant discontinuity due to thedissolution and collapse of the underlying gypsum.A similar discontinuity is reported at the end of theSalinity Crisis in other Mediterranean basins andbasin margin deposits, which evolved to brackishor freshwater Lago-Mare environments. This event,which is not synchronous in the different basins, ap-pears to represent one of the major erosional events

52 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

that occurred in the Mediterranean area over a pro-longed period of time, from the late Tortonian atleast till the end of the Messinian, and not solelyduring the period of evaporitic deposition.

Acknowledgements

This study has been funded by the EU (Hu-man Capital and Mobility Programme, ContractNo. ERBCH RXCT 930309) – Natenmar Net-work. Additional support has been provided by theBQR Programme of the Museum national d’Histoirenaturelle (1996–1997), C.I.R.I.T. Project GRQ94-1049 (University of Barcelona=CSIC) and a jointgrant from the Centre National de la RechercheScientifique=Consejo Superior de InvestigacionesCientificas. Access for sampling in the quarry Mi-nas Volcan S.A. (by kind permission of Mr. Anto-nio Miguel Ruiz-Hernadez) is gratefully acknowl-edged. Thorough reviews by B.C. Schreiber, E.Rohling (journal referees) and detailed commentsby E.K. Franseen improved the manuscript andare greatly appreciated. Begonya Illa, Jordi Illaand Adolfo Samper (Faculty of Geology, Univer-sity of Barcelona) are thanked for sample prepara-tion and analysis (B.I. was funded by the C.I.R.I.T.Project GRQ94-1049). Ramon Fontarnau from theServeis Cientıfico Tecnics (University of Barcelona)gave his advice during S.E.M. study of the sam-ples. The elemental analysis department of theCID-CSIC (Barcelona) is thanked for TOC and Sanalyses. Roser Chaler (CID-CSIC) and Pilar Teix-idor (Faculty of Geology, University of Barcelona)are thanked for performing the GC–MS analyses.Thanks are also due to P. Clement for X-ray diffrac-tion and SEM, M. Tamby and A.M. Brunet forsample processing, M. Destarac for the photographyand A. Cambreleng for the drawings. The help ofS. Servant-Vildary during the study of diatoms isgreatly acknowledged.

References

Aleria Group, 1980. Le canal de Corse et les bassins nord-tyrrheniens au Miocene superieur et terminal (Messinien); leurevolution plio-quaternaire. Geol. Mediterr. VII, 5–12.

Ayora, C., Garcıa-Veigas, J., Pueyo, J.J., 1994. X-ray microanal-ysis of fluid inclusions and its application to the geochemicalmodeling of evaporite basins. Geochim. Cosmochim. Acta 58,43–55.

Barron, J., 1992. Pliocene paleoclimatic interpretation ofD.S.D.P. Site 580 (NW Pacific) using diatoms. Mar. Micropa-leontol. 20, 23–44.

Bellanca, A., Neri, R., 1986. Evaporite carbonate cycles ofthe Messinian, Sicily: stable isotopes, mineralogy, texturalfeatures, and environmental implications. J. Sediment. Petrol.56, 614–621.

Benali, S., Schreiber, B.C., Helman, M.L., Philp, R.P., 1995.Characterization of organic matter from a restricted=evapora-tive sedimentary environment: Late Miocene of Lorca Basin,Southeastern Spain. Am. Assoc. Pet. Geol. Bull. 79, 816–830.

Benalioulhaj, S., Schreiber, B.C., Philp, R.P., 1994. Relationshipof organic geochemistry to sedimentation under highly vari-able environments, Lorca Basin (Spain): preliminary results.In: Renaut, R., Last, W. (Eds.), Sedimentology and Pale-olimnological Record of Saline Lakes. Soc. Econ. Paleontol.Mineral. Spec. Publ. 51, 315–324.

Benson, R.H., Rakic-el Bied, K., Bonaduce, G., 1991. An im-portant current reversal (influx) in the Rifian Corridor (Mo-rocco) at the Tortonian–Messinian boundary: the end of TethysOcean. Paleoceanography 6 (1), 164–192.

Bernouilli, D., Melieres, M., 1978. Dolomitization in earlyPliocene pelagic limestones, Site 374, Ionian Abyssal Plain.Init. Rep. DSDP 42A, 621–633.

Brindley, G.W., Brown, G. (Eds.), 1980. Crystal Structures ofClay Mineralogy and Their X-Ray Identification. Mineral.Soc. London, Monogr. 5.

Butler, R.W.H., Lickorish, W.H., Grasso, M., Pedley, H.M.,Ramberti, L., 1995. Tectonics and sequence stratigraphy inMessinian basins, Sicily: Constraints on the initiation and ter-mination of the Mediterranean salinity crisis. Geol. Soc. Am.Bull. 107, 425–439.

Calvert, S.E., Price, N.B., 1983. Geochemistry of Namibian Shelfsediments. In: Thiede, J., Suess, E. (Eds.), Coastal Upwelling,its Sedimentary Record, Part A: Responses of the SedimentaryRegime to Present Coastal Upwellings. Plenum Press, NewYork, pp. 337–375.

Cande, S.C., Kent, D.V., 1995. Revised calibration of the ge-omagnetic polarity timescale for the Late Cretaceous andCenozoic. J. Geophys. Res. 100, 6093–6095.

Caruso, A., Di Stefano, E., Rouchy, J.M., Sprovieri, R., 1997.Biostratigraphy, cyclostratigraphy and sedimentology controlsof the transition Tripoli-Calcare di Base (lower Messinian)from Caltanissetta Basin (Sicily). 8th Workshop of the ILPTask Force, Origin of Sedimentary Basins, Palermo, Abstr.

Cendon, D.E., Ayora, C., Garcıa-Veigas, J., Pierre, C., Pueyo,J.J., Taberner, C., 1997. The evolution of the Lorca Basin andthe composition of the Messinian ocean. EUG 9. Terra Nova 9(Suppl. 1), 412 (abstr.).

Cita, M.A., McKenzie, J.A., 1986. The terminal Miocene event.In: Hsu, K.J. (Ed.), Mesozoic and Cenozoic Oceans. Geodyn.Ser. 15, 123–140.

Cita, M.B., Wright, R.C., Ryan, W.B.F., Longinelli, A., 1978.

J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55 53

Messinian paleoenvironments. Init. Rep. DSDP 42A, 1003–1035.

Coppier, G., Ott d’Estevou, Ph., Montenat, C., 1990. Kinematicsand paleogeographic evolution of the eastern Almeria basins.Doc. Travaux IGAL, Paris 12–13, 189–194.

De Larouziere, F.D., Bolze, J., Bordet, P., Hernandez, J., Mon-tenat, C., Ott d’Estevou, P., 1988. The Betic segment ofthe lithospheric Trans-Alboran shear zone during the LateMiocene. Tectonophysics 152, 41–52.

Decima, A., Wezel, F.C., 1971. Osservazioni sulle evaporitimessiniane della Sicilia centro-meridionale. Riv. Min. Sicil.130–134, 172–187.

Decima, A., McKenzie, J.A., Schreiber, B.C., 1988. The originof ‘evaporative’ limestones: an example from the Messinian ofSicily (Italy). J. Sediment. Petrol. 58, 256–272.

Delrieu, B., Rouchy, J.M., Foucault, A., 1993. La surfaced’erosion finimessinienne en Crete centrale (Grece) et surle pourtour mediterraneen: rapports avec la crise de salinitemediterraneenne. C. R. Acad. Sci. Paris 316 II, 527–533.

Dessau, G., Jensen, M.L., Nakai, N., 1962. Geology and isotopicstudies of Sicilian sulphur deposits. Econ. Geol. 57, 410–438.

Dinares-Turell, J., Sprovieri, R., Caruso, A., Gomis-Coll, E., DiStefano, E., Pueyo, J.J., Rouchy, J.M., Taberner, C., 1998. Pre-liminary integrated magnetostratigraphic and biostratigraphiccorrelation in the Miocene Lorca Basin (Murcia, SE Spain).Acta Geol. Hisp., in press.

Docherty, C., Banda, C., 1995. Evidence for the eastward migra-tion of the Alboran Sea based on regional subsidence analysis:A case for basin formation by delamination of the subcrustallithosphere? Tectonics 14, 804–818.

Druckman, Y., Buchbinder, B., Martinotti, G.M., Siman-Tov, R.,Aharon, P., 1995. The buried Afiq Canyon (eastern Mediter-ranean, Israel): a case study of a Tertiary submarine Canyonexposed in Late Messinian times. Mar. Geol. 123, 167–185.

Fortuin, A.R., Kelling, J.M.D., Roep, T.B., 1995. The enigmaticMessinian–Pliocene section of Cuevas del Almanzora (VeraBasin, SE Spain) revisited — erosional features and strontiumisotope ages. Sediment. Geol. 97, 177–201.

Fourtanier, E., Gaudant, J., Cavallo, O., 1991. La diatomitede Castagnito (Piemont): une nouvelle preuve de l’existenced’oscillations moderees du niveau marin pendant le Messinienevaporitique. Bull. Soc. Paleontol. Ital. 30, 79–95.

Franseen, E.K., Goldstein, R.H., Farr, M.R., 1998. Quantitativecontrols on location and architecture of carbonate depositionalsequences: Upper Miocene, Cabo de Gata Region, SE Spain.J. Sediment. Res. 68, 283–298.

Fritz, P., Smith, D.G.W., 1970. The isotopic composition ofsecondary dolomites. Geochim. Cosmochim. Acta 34, 1161–1173.

Garcıa-Veigas, J., Rosell, L., Utrilla, R., Ortı, F., 1990.Aportaciones geoquımicas al conocimiento de las evapori-tas messinienses: cuencas de Palma (Mallorca), San Miguel desalinas (Alicante) y Lorca (Murcia). In: Ortı Cabo, F., SalvanyDuran, J.M. (Eds.), Formaciones Evaporıticas de la Cuenca delEbro y Cadenas Perifericas, y de la Zona de Levante. Enresa yGPPG, Universidad de Barcelona, pp. 251–256.

Garcıa-Veigas, J., Ortı, F., Rosell, L., Ingles, M., 1994. Car-

acterizacion petrologica y geoquımica de la Unidad Salinamessiniense de la cuenca de Lorca (sondeos S4 y S5). Ge-ogaceta 15, 78–81.

Garcıa-Veigas, J., Ortı, F., Rosell, L., Ayora, C., Rouchy, J.M.,Lugli, S., 1995. The Messinian salt of the Mediterranean:geochemical study of the salt from the Central Sicily Basinand comparison with the Lorca Basin (Spain). Bull. Soc. Geol.Fr. 166, 699–710.

Gaudant, J., 1989. Poissons teleosteens, bathymetrie et paleogeo-graphie du Messinien d’Espagne meridionale. Bull. Soc. Geol.Fr. V, 1161–1167.

Geel, T., 1979. Messinian gypsiferous deposits of the LorcaBasin (province of Murcia, SE Spain). Mem. Soc. Geol. Ital.1976, 26, 369–385.

Gersonde, R., Schrader, H., 1984. Marine planktic diatom cor-relation of lower-Messinian deposits in the Western Mediter-ranean. Mar. Micropaleontol. 9, 93–110.

Guillen-Mondejar, F., Rodriguez Estrella, T., Arana, R., LopezAguayo, F., 1995. Historia geologica de la cuenca de Lorca(Murcia): influencia de la tectonica en la sedimentacion. Ge-ogaceta 18, 30–33.

Hajos, M., 1986. Stratigraphy of Hungary’s Miocene Diatoma-ceous Earth Deposits. Geol. Hung., Ser. Palaeontol. 49, 209pp.

Hilgen, F.J., Krijgsman, W., Langereis, C.G., Lourens, L.J.,Santarelli, A., Zachariasse, W.J., 1995. Extending the astro-nomical (polarity) time scale into the Miocene. Earth Planet.Sci. Lett. 136, 495–510.

Howell, M.W., Thunell, R., Tappa, E., Rio, D., Sprovieri, R.,1988. Late Neogene laminated and opal-rich facies from theMediterranean region: geochemical evidence for mechanismsof formation. Palaeogeogr., Palaeoclimatol., Palaeoecol. 64,265–286.

Hsu, K.J., Montadert, L., Bernouilli, D., Cita, M.B., Erikson, A.,Garrison, R.G., Kidd, R.B., Melieres, F., Muller, C., Wright,R., 1978. History of the Mediterranean salinity crisis. Init.Rep. DSDP 42A, 1053–1078.

IGME, 1982. Ampliacion de la investigacion de pizarras bitu-minosas en la zona de lorca (Murcia) (Fase II): InscripcionesLorca 134 (Murcia) y Lorca bis 155 (Murcia). Inst. Geol.(Min.) Espana, Madrid, 116 pp.

Imbrie, J., Kipp, N.G., 1971. A new micropaleontologicalmethod for quantitative paleoclimatology: Application to aLate Pleistocene Caribbean core. In: Turekian, K.K. (Ed.),The Late Cenozoic Glacial Age. Yale University Press, NewHaven, CT, pp. 71–81.

Jouse, O.G., Kozlova, G., Muhina, V., 1971. Distribution of di-atoms in sediment from the Pacific. In: Funnell, B.M., Riedel,W.R. (Eds.), The Micropaleontology of Oceans. CambridgeUniversity Press, Cambridge, 828 pp.

Keigwin, L.D., Aubry, M.P., Kent, D.V., 1987. North Atlanticlate Miocene stable isotope stratigraphy, biostratigraphy andmagnetostratigraphy. Init. Rep. DSDP 94, 935–963.

Klovan, J.E., Imbrie, J., 1971. An algorithm and Fortran IVprogram for large scale Q-mode factor analysis. J. Int. Assoc.Math. Geol. 3, 61–77.

54 J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55

Koizumi, I., 1975. Neogene diatoms from the NorthwesternPacific Ocean. Init. Rep. DSDP 32, 865–880.

Krijgsman, W., Hilgen, F.J., Langereis, C.G., Santarelli, A.,Zachariasse, W.J., 1995. Late Miocene magnetostratigraphy,biostratigraphy and cyclostratigraphy in the Mediterranean.Earth Planet. Sci. Lett. 136, 475–494.

Mansour, B., Moissette, P., Noel, D., Rouchy, J.M., 1995.L’enregistrement par les associations de diatomees des envi-ronnements messiniens: l’exemple de la coupe de Sig (Bassindu Chelif-Algerie). Geobios 28, 261–279.

Matthews, A., Katz, A., 1977. Oxygen isotope fractionationduring the dolomitization of calcium carbonate. Geochim.Cosmochim. Acta 41, 1431–1438.

McKenzie, J.M., 1985. Stable isotope mapping in Messinianevaporative carbonates of Central Sicily. Geology 13, 851–854.

McKenzie, J.A., Jenkyns, H.C., Bennett, G.G., 1979. Stableisotope study of the cyclic diatomite–claystones from theTripoli formation, Sicily: a prelude to the Messinian salinitycrisis. Palaeogeogr., Palaeoclimatol., Palaeoecol. 29, 125–142.

Montenat, C., 1973. Les formations neogenes et quaternaires duLevant espagnol (Provinces d’Alicante et de Murcie). TheseSci., Univ. Orsay, 1170 pp.

Montenat, C., 1977. Les bassins neogenes du Levant d’Alicanteet de Murcia (Cordilleres betiques orientales, Espagne). Strati-graphie, paleogeographie et evolution dynamique. Doc. Lab.Geol. Fac. Sci. Lyon 69, 345 pp.

Montenat, C., Ott d’Estevou, Ph., Masse, P., 1987. Tectonic–sedimentary characters of the Betic Neogene basins evolvingin a crustal transcurrent shear zone (SE Spain). Bull. Cent.Rech. Explor.-Prod. Elf-Aquitaine 11, 1–22.

Montenat, C., Ott d’Estevou, P., Delort, T., 1990. Le bassin deLorca. Doc. Travaux IGAL Paris 12–13, 261–280.

Muller, D.W., Hsu, K.J., 1987. Event stratigraphy and paleo-ceanography in the Fortuna basin (southeast Spain): A sce-nario for the Messinian salinity crisis. Paleoceanography 6,679–696.

Ogniben, L., 1957. Petrografia della Serie Solfifera Siciliana econsiderazioni geologiche relative. Mem. Descrit. Carta Geol.Ital. 33, 275 pp.

Ortı, F., 1990. Introduccion a las evaporitas de la Cuenca deLorca. In: Ortı Cabo, F., Salvany Duran, J.M. (Eds.), For-maciones evaporıticas de la Cuenca del Ebro y cadenas per-ifericas, y de la zona de Levante. Nueva aportaciones y guıade superficie. Enresa y GPPG, Universidad de Barcelona, pp.257–266.

Ortı, F., Rosell, L., 1990. Yesos de Lorca. In: Ortı Cabo, F.,Salvany Duran, J.M. (Eds.), Formaciones evaporıticas de laCuenca del Ebro y cadenas perifericas, y de la zona deLevante. Nueva aportaciones y guıa de superficie. Enresa yGPPG, Universidad de Barcelona, pp. 297–298.

Ortı, F., Garcıa-Veigas, J., Rosell, L., Rouchy, J.M., Ingles,M., Gimeno, D., Kasprzyk, A., Playa, E., 1993. Correlacionlitoestratigrafica de las evaporitas mesinienses en las cuencasde Lorca y Fortuna (Murcia). Geogaceta 14, 98–101.

Parker, F.L., 1958. Eastern Mediterranean foraminifera. Swed.Deep-Sea Exped. Rep. 8, 219–283.

Permanyer, A., Baranger, R., Lugardon, B., 1994. Caracterisa-tion des schistes bitumineux des sediments pre-evaporitiquesmessiniens du bassin de Lorca (sud-est de l’Espagne). Bull.Cent. Rech. Explor.-Prod. Elf Aquitaine 18, 135–149.

Perrodon, A., 1957. Etude geologique des bassins neogenessub-littoraux de l’Algerie occidentale. Bull. Serv. Carte Geol.Alger. (h.s) 12, 328 pp.

Pierre, C., 1982. Teneurs en isotopes stables (18O, 2H, 13C, 34S)et conditions de genese des evaporites marines: application aquelques milieux actuels et au Messinien de la Mediterranee.These, Univ. Paris-Sud, Orsay, 266 pp.

Pierre, C., Rouchy, J.M., 1988. The carbonate replacementsafter sulfate evaporites in the Middle Miocene of Egypt. J.Sediment. Petrol. 58, 446–456.

Pierre, C., Blanc-Valleron, M.M., Caulet, J.P., Cespuglio, G.,Rouchy, J.M., Sprovieri, R., Di Stefano, E., Caruso, A.,Taberner, C., Pueyo, J.J., 1997. Sedimentary, stable iso-topes and micropaleontological records of paleoceanographicchanges during the deposition of the lower Messinian TripoliFormation of the Gibliscemi-Falconara section. InternationalConference on Neogene Paleoceanography, Erice, Abstr. Vol.,pp. 54–55.

Platt, J.P., Vissers, R.L.M., 1989. Extensional collapse of thick-ened continental lithosphere: A working hypothesis for theAlboran Sea and Gibraltar Arc. Geology 17, 540–543.

Rouchy, J.M., 1976. Sur la genese de deux principaux types degypse (finement lite et en chevrons) du Miocene terminal deSicile et d’Espagne meridionale. Rev. Geogr. Phys. Geol. Dyn.XVIII, 347–364.

Rouchy, J.M., 1982. La genese des evaporites messiniennes deMediterranee. Mem. Mus. Natl. Hist. Nat. L, 267 pp.

Rouchy, J.M., 1986. La formation des grandes accumulationsevaporitiques d’origine marine: les enseignements tires del’etude des series miocenes de la Mediterranee et de la MerRouge. Bull. Soc. Geol. Fr. 3, 511–520.

Rouchy, J.M., Pierre, C., 1979. Donnees sedimentologiques etisotopiques sur les gypses des series evaporitiques messini-ennes d’Espagne meridionale et de Chypre. Rev. Geogr. Phys.Geol. Dyn. 21, 267–280.

Rouchy, J.M., Pierre, C., 1987. Authigenic natroalunite in theMiddle Miocene evaporites from the Red Sea (Guemsah,Egypt). Sedimentology 34, 807–812.

Rouchy, J.M., Saint-Martin, J.P., 1992. Late Miocene events inthe Mediterranean as recorded by carbonate–evaporite rela-tions. Geology 20, 629–632.

Russell, M., Grimalt, J.O., Hartgers, W.A., Taberner, C., Rouchy,J.M., 1997. Bacterial and algal markers in sedimentary or-ganic matter deposited under natural sulphurization conditions(Lorca Basin, Murcia, Spain). Org. Geochem. 26, 605–625.

Santisteban, C., 1981. Petrologıa y sedimentologıa de los materi-ales del Mioceno superior de la Cuenca de Fortuna (Murcia) ala luz de la teorıa de la crisis de salinidad. Thesis Doct., Univ.Barcelona, 725 pp.

Sanz de Galdeano, C., 1990. Geological evolution of the BeticCordilleras in the Western Mediterranean, Miocene to present.Tectonophysics 172, 107–119.

J.M. Rouchy et al. / Sedimentary Geology 121 (1998) 23–55 55

Schrader, H.J., Gersonde, R., 1978a. Diatoms and silicoflagel-lates. Utrecht Micropalaeontol. Bull. 17, 129–176.

Schrader, H.J., Gersonde, R., 1978b. The late Messinian Mediter-ranean brackish to freshwater environment. Diatom floral evi-dence. Init. Rep. DSDP 42A, 761–769.

Schreiber, B.C., 1974. Upper Miocene (Messinian) EvaporiteDeposits of the Mediterranean Basin and Their DepositionalEnvironments. Thesis, Rensselaer Polytech. Inst., New York,395 pp.

Sen Gupta, B.K., Machain-Castillo, M.L., 1993. Benthicforaminifera in oxygen-poor habitas. Mar. Micropaleontol. 20,183–201.

Sharma, T., Clayton, R.N., 1965. Measurements of 18O=16Oratios of total oxygen and carbonates. Geochim. Cosmochim.Acta 29, 1347–1353.

Shearman, D.J., 1978. Evaporites of coastal sabkhas. In: Dean,W.E., Schreiber, B.C. (Eds.), Marine Evaporites. Soc. Econ.Paleontol. Mineral. Short Course 4, 6–42.

Sinninghe Damste, J.S., de Leeuw, J.W., 1987. The origin andfate of C-20 and C-15 isoprenoid sulphur compounds in sedi-ments and oils. Int. J. Environ. Anal. Chem. 28, 1–19.

Sinninghe Damste, J.S., Rijpstra, W.I.C., de Leeuw, J.W.,Schenck, P.A., 1989. The occurrence and identification ofseries of organic sulphur compounds in oils and sedimentextracts. II. Their presence in samples from hypersalineand non-hypersaline paleoenvironments and possible applica-tions as source, palaeoenvironmental and maturity indicators.Geochim. Cosmochim. Acta 53, 1323–1341.

Sprovieri, R., Di Stefano, E., Sprovieri, M., 1996. High resolu-tion chronology for late Miocene Mediterranean stratigraphicevents. Riv. Ital. Paleontol. Stratigr. 102, 77–104.

Sprovieri, R., Di Stefano, E., Sprovieri, M., 1997. Integratedcalcareous plankton biostratigraphy and cyclostratigraphy inthe Licata Tripoli Formation in Sicily. 8th Workshop of theILP Task Force, Origin of Sedimentary Basins, Palermo, pp.114–115.

Teixidor, P., Grimalt, J.O., Pueyo, J.J., Rodriguez-Valera, F.,1993. Isopranylglycerol diethers in non-alkaline evaporitic en-vironments. Geochim. Cosmochim. Acta 57, 4479–4489.

Utrilla, R., Pierre, C., Ortı, F., Pueyo, J.J., 1992. Oxygen andsulphur isotope compositions as indicators of the origin ofMesozoic and Cenozoic evaporites from Spain. Chem. Geol.(Isot. Geosci. Sect.) 102, 229–244.

Van der Zwaan, G.J., 1979. The pre-evaporite Late Miocene en-vironment of the Mediterranean: stable isotopes of planktonicforaminifera from section Falconara, Sicily. K. Ned. Akad.Wetensch. 82, 487–502.

Verhallen, P.J.J.M., 1991. Late Pliocene to early PleistoceneMediterranean mud-dwelling foraminifera; influence of achanging environment on community structure and evolution.Utrecht Micropalaeontol. Bull. 40, 187 pp.

Volkman, J.K., Barret, J.M., Dunstan, G.A., 1994. C25 and C30

highly branched isoprenoid alkenes in laboratory cultures oftwo marine diatoms. Org. Geochem. 21, 407–413.

Wright, R., 1978. Neogene paleobathymetry of the Mediter-ranean based on benthic foraminifers from DSDP Leg 42A.Init. Rep. DSDP 42A, 837–846.

Zhang, J., Scott, D.B., 1996. Messinian deep-water turbidites andglacioeustatic sea-level changes in the North Atlantic: Linkageto the Mediterranean Salinity Crisis. Paleoceanography 11,277–297.