repetitive changes in early pliocene vegetation revealed ...€¦ · ent for north-european to...

Repetitive changes in Early Pliocene vegetation revealedby high-resolution pollen analysis :

revised cyclostratigraphy of southwestern Romania

Speranta-Maria Popescu �

Laboratoire ‘Pale¤oEnvironnements et Pale¤obioSphe're’, Universite¤ Claude Bernard - Lyon 1, 27-43 boulevard du 11 Novembre,69622 Villeurbanne Cedex, France

Received 16 February 2001; received in revised form 11 October 2001; accepted 29 October 2001

Abstract

A high-resolution pollen analysis has been carried out on the Lupoaia section (SW Romania) in order to checkwhether the repetitive clay^lignite alternations correspond to cyclic changes in climate. Increases in altitudinal treepollen content appear to have been caused by drops in temperature, while developments of thermophilous elementscorrespond to rises in temperature, still under humid conditions. Such repeated changes in vegetation, on the wholeconsistent with the clay^lignite alternations, have been forced by cycles in eccentricity. On the basis of a comparisonbetween the Lupoaia pollen record and (1) European climatostratigraphy (based on reference pollen diagramsdocumenting global changes), and (2) global climatic curves (eccentricity, N18O), the age of the section has beenreconsidered. The Lupoaia section (i.e. from lignite IV to lignite XIII) starts just before the C3n.3n Chron andprobably ends just before the C3n.1n Chron. The section represents a time span of about 600 kyr, i.e. from about 4.90Ma to about 4.30 Ma. 7 2001 Elsevier Science B.V. All rights reserved.

Keywords: pollen analysis; vegetation; cyclostratigraphy; Pliocene; Romania

1. Introduction

During the last decades, palynological e¡ortshave been made to obtain a good knowledge ofthe European and Mediterranean Pliocene vegeta-tion and climate (Zagwijn, 1960; Menke, 1975;Diniz, 1984; Suc, 1984a; Zheng and Cravatte,1986; Drivaliari, 1993; Bertini, 1994). Today, re-constructions of vegetation and climate are coher-ent for North-European to Mediterranean lati-

tudes during the whole Pliocene, more especiallythe Early Pliocene (i.e. the Zanclean stage from5.32 to 3.6 Ma) (Suc and Zagwijn, 1983; Suc etal., 1995). In addition, a climate transfer function,based on pollen records, has been introduced,which underlines subtropical temperatures insouthwestern Europe during the Early Pliocene(for example, 16.5‡C as mean annual temperaturein the Nice area) (Fauquette et al., 1999a).

Early Pliocene pollen data show consistent veg-etation changes in Europe from north (Zagwijn,1960; Menke, 1975) to south (Suc, 1984a; Bertini,1994; Zheng and Cravatte, 1986), as well as fromwest (Diniz, 1984) to east (Drivaliari et al., 1999),

0034-6667 / 01 / $ ^ see front matter 7 2001 Elsevier Science B.V. All rights reserved.PII: S 0 0 3 4 - 6 6 6 7 ( 0 1 ) 0 0 1 4 2 - 7

* Fax: +33-4-72448382.E-mail address: [email protected] (S.-M. Popescu).

PALBO 2442 6-8-02

Review of Palaeobotany and Palynology 120 (2001) 181^202

www.elsevier.com/locate/revpalbo

mailto:[email protected]

http://www.elsevier.com/locate/revpalbo

that allow the subdivision of the period into threemain climatic phases. Two main warm phases(Brunssumian A and C of Zagwijn (1960), pollenzones P Ia and P Ic of Suc (1984a)) encompassthe ¢rst relative cooling episode (Brunssumian Bof Zagwijn (1960), pollen zone P Ib of Suc(1984a)). This subdivision is in accordance withthe large variations observed in the N

18O curvefrom the ocean record (Tiedemann et al., 1994;Shackleton et al., 1995). Most of the Europeanreference pollen diagrams include secondary £uc-tuations within the above mentioned climaticphases of the Early Pliocene (Susteren in TheNetherlands: Zagwijn, 1960; Oldenswort 9 inGermany: Menke, 1975; Garraf 1 in the north-western Mediterranean shelf o¡-shore Spain: Sucand Cravatte, 1982; Rio Maior F 14 in Portugal:Diniz, 1984). Unfortunately, these pollen recordshave been established on a somewhat low numberof samples, that do not allow accurate correlationwith the continuous N

18O record, and thereforewith the astronomic parameters.

Van der Zwaan and Gudjonsson (1986) haveestablished the ¢rst N

18O record at a relativehigh-resolution for the Pliocene of Sicily which

shows many £uctuations. More recently, Tiede-mann et al. (1994) and Shackleton et al. (1995)have demonstrated that such N

18O £uctuationsare astronomically forced by precession (ca. 20-kyr cycles). In addition, Hilgen (1990, 1991a,b)has established a very accurate time scale for theMediterranean Lower Pliocene based both onmagnetostratigraphy and sedimentological cyclo-stratigraphy. It was demonstrated that rapidchanges in sedimentation were simultaneouslyforced by precession and eccentricity (100- and400-kyr cycles).

Certain long, continuous and very propitiousland sections can provide high-resolution pollenrecords that may reveal whether the changes invegetation are induced by astronomic cycles.The pollen records from Stirone and Monticino(Lower Zanclean of the Po Valley) suggest, de-spite a low sampling resolution, cyclic variationsin the vegetation that are forced by eccentricity(Bertini, 1994). The Lupoaia section (southwest-ern Romania; Fig. 1) o¡ers a long Early Pliocenesedimentary record, characterised by almost reg-ular clay^lignite alternations which have beenlinked to eccentricity forcing (Van Vugt et al.,

Fig. 1. Location of the Lupoaia section in the Dacic Basin. (1) Lupoaia, (2) Ticleni borehole.

PALBO 2442 6-8-02

S.-M. Popescu / Review of Palaeobotany and Palynology 120 (2001) 181^202182

2001). However, Van Vugt et al. (1998) have dem-onstrated that the lignite^clay alternations of theEarly Pliocene Ptolemais lacustrine basin (north-western Greece) were forced by precession: theselignites would correspond to drier phases andclays to moister phases (Kloosterboer-van Hoeve,2000).

Accordingly, a high-resolution pollen investiga-tion was carried out on the Lupoaia long sectionin order (1) to know whether Early Pliocene veg-etation changes are related to astronomical cycles,and (2) to contribute to the clari¢cation of theclimatic signi¢cance of lignites located in the east-ern Mediterranean region.

2. The Lupoaia section

2.1. General characters

The Lupoaia quarry is located near the city ofMotru (district of Gorj, Romania) close to theCarpathians in Oltenia, some 30 km north ofthe Danube river (Fig. 1). The quarry is especiallycharacterised by clay^lignite alternations, and in-cludes several sand layers of various thickness.The sedimentary record toward the base of theseries has been recovered by cored boreholes.

The Lupoaia quarry is 121.50 m high andshows nine major lignite beds corresponding tolignite V to lignite XIII in the regional nomencla-ture. Some of them have been subdivided into twoor three secondary layers as lignites VIII and X(Fig. 2). The underlying sediments cored in bore-holes F6 (thickness: 21.20 m) and F11 include thelower layer of lignite V and lignite IV. The presentpalynological study begins below lignite IV and¢nishes above lignite XIII (Fig. 2), correspondingto a thickness of 133.70 m. Some other thin ligniteor lignitic clay layers have been recorded. Blue-grey clays are abundant in the section and rich inleaf prints (Ticleanu and Buliga, 1992; Ticleanuand Diaconita, 1997). Fluvial sands are mainlyconcentrated in the mid and the upper part ofthe section (Fig. 2). The thickness of the di¡erentlayers varies within the Motru Basin; for exam-ple, sands are thicker in the south and the lignitebeds contain more subdivisions towards the

south-east. According to Ticleanu and Diaconita(1997), the Motru lignites belong to a deltaic sys-tem which, according to Clauzon and Suc (per-sonal information), £owed into the Dacic Seaclose to the Zanclean Danube delta, in the areaof Dobreta Turnu Severin.

2.2. Chronology

Radan and Radan (1998) have performed pa-laeomagnetic measurements of samples from boththe Lupoaia quarry (more than 1000 samples) andtwo cored boreholes F6 and F11 (about 380 sam-ples). Two normal palaeomagnetic events wereidenti¢ed in the lowermost part (from clays over-lying lignite IV to the upper lignite V) and in themid-part of the section (from about lignite VII tolignite VIII; Fig. 2), respectively. The lower nor-mal episode is preceded by a relatively prolongedundetermined polarity zone, while another onehas been identi¢ed above lignite XIII (Fig. 2).Some other thin undetermined polarity zones oc-cur along the section but they mostly correspondto lignites which do not constitute a very goodlithology for palaeomagnetic measurements. Ra-dan and Radan (1998) have considered these nor-mal episodes to represent the C3n.2n Chron (i.e.the Nunivak Chron) and the C3n.1n Chron (i.e.the Cochiti Chron), respectively. To identify thetwo normal episodes, the authors used as bio-stratigraphic argument the presence within ligniteVIII (Fig. 2) of a primitive Mimomys (Mimomysrhabonensis) considered to be representative of thelowermost mammal zone MN 15 (Radulescu etal., 1997). However, it is assumed that the prim-itive Mimomys have appeared in Romania at thesame time as in the northwestern Mediterraneanregion, i.e. after the C3n.1n Chron (locality ofMas Soulet, N|“mes area) (Aguilar and Michaux,1984; Aguilar et al., 1999). In other words, it isconsidered that the ‘boundary’ between mammalzones MN 14 and MN 15 is coeval in easternEurope and western Europe, a concept that iscurrently topic of strong international discussion(ad hoc Working Group of the Regional Commit-tee on the Mediterranean Neogene Stratigraphy).In northern Greece, the ¢rst Mimomys (Mimomysdavakosi) is recorded in the Ptolemais 3 locality

PALBO 2442 6-8-02

S.-M. Popescu / Review of Palaeobotany and Palynology 120 (2001) 181^202 183

Fig. 2.

PALBO 2442 6-8-02


(Van de Weerd, 1978) that would belong to theC3n.3n Chron (Van Vugt et al., 1998) accordingto the stratigraphic information provided by Vande Weerd (1978), Koufos and Pavlides (1988) andVan Vugt et al. (1998). Magnetostratigraphy ofthe Ptolemais section suggests that Mimomys ap-peared earlier in eastern Europe than in westernEurope. This view is also supported by the ¢rstappearance of the ancestor of Mimomys, Promi-momys insuliferus, which, in the Ptolemais section(Kardia locality; Van de Weerd, 1978) would oc-cur in a time span including the C3n.4n Chron(Van Vugt et al., 1998) according to the localstratigraphic information (Van de Weerd, 1978;Koufos and Pavlides, 1988; Van Vugt et al.,1998). As a consequence, there is no reliable argu-ment provided by micromammals to support thechronological assignment of the Lupoaia sectionto the successive chrons C3n.2n and C3n.1n. Us-ing a cryogenic magnetometer, Van Vugt et al.(2001) have con¢rmed the polarity reversals ofthe Lupoaia quarry (78 samples going from ligniteV to lignite IX), including two relatively pro-longed undetermined polarity zones (within andabove lignite V, within lignite VII) (Fig. 2). Theauthors followed the same chronological interpre-tation; in their opinion, the reversed episode re-corded in the upper part of the section could beassigned, considering its relatively important du-ration, to the long reversal C2Ar. Such an inter-pretation does not take into account possiblechanges in sedimentation rate in the uppermostpart of the section, which is richer in lignite andsand beds.

Among the other micromammal remains foundin the lignite bed VIII of the quarry, there is Apo-demus dominans, which is a common species in theEuropean Pliocene, and also found in the threemammal localities of Ptolemais (Van de Weerd,1978).

Further information is provided by the discov-ery of Dicerorhinus megarhinus remains in the

HorasRti mine (Motru Basin) lignite X (Apostoland Enache, 1979). According to Gue¤rin (1980),this species corresponds to the Lower Pliocene(Mammal Zones 14 and 15). The size of the Ho-rasRti mine specimen (Apostol and Enache, 1979)is almost similar to that from the specimen ofMontpellier (De Serres, 1819). In western Europe,the range of the species should be comprised be-tween about 5 and 4 Ma.

Therefore, mammal fauna does not provide anunquestionable chronological support to the pro-posed magnetostratigraphic assignment (Radanand Radan, 1998; Van Vugt et al., 2001); onthe contrary, the data suggest an age older thanthat proposed for the Lupoaia section.

3. Materials and methods

Numerous samples (231) of clays and ligniteshave been analysed from the Lupoaia section.Most of them (204) have provided a rich pollen£ora. Samples 1^31 come from the cored boreholeF6, samples 32^204 come from the quarry (Fig.2).

Clay samples have been prepared following theclassic method (successive actions by HCl, HF,etc. ; concentration of palynomorphs using ZnCl2at density = 2, then ¢ltering at 10 Wm). Anothertechnical approach has been used for the lignites,starting with a KOH action which replaces theacids. Samples are mounted in glycerine to allowcomplete observation of palynomorphs, necessaryfor their proper botanical identi¢cation. Morethan 150 pollen grains (Pinus, a generally over-represented element in the pollen £ora, excluded)have been counted per sample. Results are givenin a simpli¢ed detailed pollen diagram where pol-len percentages are calculated relative to the totalpollen sum (Fig. 3). Spores (pteridophytes, bryo-phytes, fungi), algae and dino£agellate cysts havebeen scored separately. Dry samples have been

Fig. 2. Lithological succession of the Lupoaia section (quarry and cored borehole F6) and its magnetostratigraphic subdivision(Radan and Radan, 1998; Van Vugt et al., 2001); stratigraphic locations of mammal remains are indicated. Lithology: (1) lig-nite; (2) lignitic clay; (3) clay; (4) sand; (5) sandy clay; (6) not recovered interval within borehole F6.

PALBO 2442 6-8-02


weighted, and volumes of residues after treatmentwere measured in order to calculate pollen con-centrations using the method described by Cour(1974). A synthetic pollen diagram has been con-structed according to the ecological requirementsof taxa (Fig. 4). Such standard synthetic pollendiagrams (Suc, 1984a) clearly document temporalchanges in pollen content, and aid in comparisonsbetween pollen records throughout Europe andthe Mediterranean region; they are generallyused for detecting long-distance climatostrati-graphic relationships (Suc et al., 1995).

Interpretations of the pollen records are alsosupported by statistical analyses (principle com-ponent analysis, spectral analysis).

Pollen and spore counts from the Lupoaia sec-tion are archived at the Laboratory ‘Pale¤oEnvi-ronnements et Pale¤obioSphe're’ (UniversityClaude Bernard - Lyon 1) and will be available

from the ‘Cenozoic Pollen and Climatic values’database (CPC)1.

4. Pollen £ora and vegetation

The pollen £ora is very rich (137 taxa; Fig. 3),and dominated by trees. These results consider-ably increase the known £oristic palaeodiversityof the area (the delta environment to the altitudi-nal belts of the Carpathians), when compared tothe preliminary palynological study of Petrescu etal. (1989), and the analyses of palaeobotanicalmacrofossils (Ticleanu, 1992; Ticleanu and Dia-

Fig. 3.

1 For further information, contact Dr. Se¤verine Fauquette(same address as the author), curator of the CPC database([email protected]).

PALBO 2442 6-8-02


conita, 1997). In general, pollen concentration isrelatively high (more than 1000 pollen grains/g ofsediment), but shows important variations (from50 up to 83 000 pollen grains/g of sediment).

The dominant taxa have been used to describeand interpret the detailed pollen diagrams interms of the following vegetational units:

(1) Subtropical swamp forests with the Taxo-

Fig. 3. Simpli¢ed detailed pollen diagram of the Lupoaia section. Some scarcely represented taxa are grouped into the followingsections: (1) megathermic elements: Euphorbiaceae p.p., Amanoa, Meliaceae, Entada type, other Mimosoideae, Pachysandra type,Sapindaceae, Loranthaceae, Tiliaceae p.p. ; (2) mega-mesothermic swamp elements: Cyrillaceae^Clethraceae, Myrica, Taxodiumtype, Cephalanthus, Nyssa (Nyssa cf. sinensis and Nyssa cf. aquatica) ; (3) other Taxodiaceae: Sequoia type, non-identi¢ed Taxo-diaceae; (4) other mega-mesothermic elements: Arecaceae, Sapotaceae, Anacardiaceae, Araliaceae, Microtropis fallax, Distyliumcf. sinensis, Parrotiopsis cf. jacquemontiana, Leea, Magnolia ; (5) other Ulmaceae: Celtis, Ulmus, Ulmus^Zelkova type; (6) othermesothermic elements (1): Carpinus cf. orientalis, Platanus, Ostrya, Liquidambar (including Liquidambar cf. orinetalis) ; (7) othermesothermic elements (2): Vitis, Hedera (including Hedera cf. helix), Lonicera, Buxus sempervirens type, Ligustrum ; (8) other me-sothermic elements (3): Sambucus, Viburnum, Rhus, Tilia, Ilex, Acer, Tamarix ; (9) other Carpinus : Carpinus cf. betulus, non-iden-ti¢ed Carpinus ; (10) riparian trees: Alnus, Salix, Populus, Fraxinus ; (11) Mediterranean xerophytes: Olea, Phillyrea, Pistacia,Quercus ilex type, Cistus, Periploca, Phlomis cf. fruticosa ; (12) subdesertic elements: Nolina, Prosopis ; (13) other Asteraceae: As-teroideae, Cichorioideae, Centaurea ; (14) other halophytes: Ephedra, Caryophyllaceae, Plumbaginaceae; (15) other herbs: Erodi-um, Convolvulus, Linum, Mercurialis, Euphorbia, Brassicaceae, Apiaceae, Scabiosa, Knautia, Malvaceae, Borraginaceae, Helianthe-mum, Plantago, Rumex, Polygonum (including Polygonum cf. aviculare and Polygonum cf. lapatifolium), Rosaceae, Asphodelus,other Liliaceae, Cannabis, other Cannabaceae, Papillionideae; (16) freshwater plants: Thalictrum, other Ranunculaceae, Butomus,Restio, other Restionaceae, Myriophyllum, Potamogeton, Sparganium, Typha, Nuphar, Nymphaea, Oenotheraceae, Trapa, Utricu-laria, other Monocotyledones; (17) non-identi¢ed pollen grains include Gymnocardiidites subrotunda and Tricolporopollenites sibiri-cum. Percentages are calculated relative to the total pollen sum.

PALBO 2442 6-8-02


PALBO 2442 6-8-02


diaceae as main component, in which regular oc-currences of Taxodium pollen-type are found, in-cluding Taxodium and Glyptostrobus. Today,Taxodium distichum is found in the coastalswamps of Northeastern America (Florida andMississippi delta; George, 1972; Roberts, 1986),and Glyptostrobus grows in swampy lowlandswithin the evergreen broad-leaved forest in China(Wang, 1961). According to macrofossil evidence,the Taxodiaceae living in the Lupoaia Plioceneswamps belong to Glyptostrobus (Glyptostrobuseuropaeus ; Ticleanu, 1992). Cupressaceae pollengrains are also abundant but cannot be identi¢edmore accurately. As their frequency trends resem-ble those of the Taxodiaceae, it is assumed thatmost Cupressaceae were represented by taxa re-quiring warm and humid conditions, such asthose living today in subtropical Asia (Chamaecy-paris). This is also supported by the position ofCupressaceae on the second axis of the principlecomponent analysis (Fig. 5). Some other elementsformed part of these associations, includingCephalanthus, Myrica or Nyssa cf. sinensis. In ad-dition, these assemblages probably containedsome subtropical to warm-temperate elements ;Salix p.p., Alnus p.p., Populus p.p. and a Juglansspecies which is morphologically similar to Ju-

glans cathayensis, were found among the plantmacrofossils (Ticleanu, personal information).These elements may be the ¢rst indicators of theforthcoming warm-temperate swamps of the Mid-dle-Late Pliocene (Drivaliari et al., 1999). Severalpteridophytes including Osmunda were found inthe swamp assemblages.

(2) The abundance of Cyperaceae pollen sug-gests the presence of herbaceous marshes (seethe result of the principle component analysis ;Fig. 5). Water plants were also present in suchan environment (Restionaceae, Myriophyllum, Po-tamogeton, Trapa, Typha, Sparganium, Nuphar,Oenotheraceae). These two pollen assemblageshave a strong resemblance to the modern vegeta-tion of South Florida (George, 1972) and the Mis-sissippi delta (Roberts, 1986), where swamp for-ests (with Taxodium distichum) and herbaceousmarshes (with Cyperaceae mainly) coexist. Theirrespective £oristic composition can also be seen inthe modern surface pollen spectra (Rich, 1985;Suc, personal information). Their constitution isin agreement with macro£ora (Ticleanu, 1992; Ti-cleanu and Diaconita, 1997).

(3) Behind such a coastal vegetation, and up tomid-altitude, a mixed subtropical to warm-tem-perate forest existed with mega-mesothermic and

Fig. 4. (A) Synthetic pollen diagram of the Lupoaia section. Pollen grains have been grouped according to ecological signi¢canceof the concerned plants. Lithology: see Fig. 2. Pollen assemblages: (1) megathermic elements (unidenti¢ed Euphorbiaceae, Ama-noa, Mimosaceae including Entada and Pachysandra types, Meliaceae, Sapindaceae, Loranthaceae, Arecaceae, Sapotaceae, Tilia-ceae); (2) mega-mesothermic elements (mainly Taxodiaceae, Engelhardia, Cephalanthus, Distylium, Parrotiopsis jacquemontiana,Microtropis fallax, Cyrillaceae^Clethraceae, Leea, Myrica, Nyssa sinensis, Parthenocissus henryana, Ilex £oribunda type, Anacar-diaceae, Araliaceae, Magnolia) ; (3) lower^mid-altitude coniferous elements, Cathaya ; (4) mesothermic elements (deciduous Quer-cus chie£y, Carya, Pterocarya, Liquidambar, Parrotia persica, Carpinus, Ulmus, Zelkova, Celtis, Ostrya, Platanus, Juglans, Juglanscf. cathayensis, Nyssa, Sciadopitys, Buxus sempervirens type, Acer, Tilia, Fagus, Alnus, Salix, Populus, Ericaceae, Vitis, Hedera,Lonicera, Fraxinus, Ligustrum, Sambucus, Viburnum, Rhus, Ilex, Tamarix, Betula) ; (5) Pinus. Meso-microthermic trees: (6) mid-al-titude trees, Cedrus, Keteleeria and Tsuga ; (7) montane trees, Abies and Picea ; (8) elements without signi¢cance (Rosaceae, Ra-nunculaceae, unidenti¢ed pollen grains, poorly preserved pollen grains); (9) Cupressaceae; (10) ‘Mediterranean’ xerophytes suchas Olea, Phillyrea, Quercus ilex type, Pistacia, Cistus, Phlomis cf. fruticosa, Periploca. Herbs: (11) Cyperaceae, Poaceae, Astera-ceae, Plantago, Brassicaceae, Apiaceae, Polygonum, Rumex, Amaranthaceae^Chenopodiaceae, Caryophyllaceae, Linum, Erodium,Convolvulus, Mercurialis, Euphorbia, Scabiosa, Knautia, Malvaceae, Borraginaceae, Helianthemum, Asphodelus, Liliaceae, Cannaba-ceae, Fabaceae, Plumbaginaceae, Butomus, water plants such as Potamogeton, Restionaceae, Myriophyllum, Typha, Sparganium,Thalictrum, Nuphar, Nymphaea, Oenotheraceae, Trapa, Utricularia ; (12) steppe elements (Artemisia, Ephedra). (B) Record of ther-mophilous trees (i.e. the megathermic elements (such as Mimosaceae including Proposis and Entada type, Pachysandra type, Me-liaceae, Euphorbiaceae including Amanoa, Sapindaceae, Loranthaceae, Tiliaceae, Arecaceae, Sapotaceae) and the mega-mesother-mic elements such as Taxodiaceae (Sciadopitys excluded), Juglans cf. cathayensis, Engelhardia, Myrica, Nyssa cf. sinensis,Distylium, Parthenocissus cf. henryana, Parrotiopsis cf. jacquemontiana, Microtropis fallax, Ilex £oribunda type, Cyrillaceae^Cleth-raceae, Leea, Magnolia, Nolina, plus Cupressaceae] in contrast to altitudinal trees (Tsuga, Cedrus, Cathaya, Abies and Picea)along the Lupoaia section.

PALBO 2442 6-8-02


mesothermic elements such as Engelhardia, Carya,Pterocarya, deciduous Quercus, Fagus, some Cu-pressaceae, Juglans, Zelkova, Carpinus, Acer, etc.Here, the riparian associations were richer in Al-nus p.p., Salix p.p., Liquidambar, Parrotia andPopulus p.p. Some other Taxodiaceae, such asSciadopitys and those corresponding to the Se-quoia pollen type, are represented by a smallquantity of pollen grains. As considered by Ti-cleanu and Diaconita (1997), these elements areincluded in this vegetation group, taking into ac-count their modern habitat which is often altitu-dinal and in a lower latitudinal range than inEurope, and because they occur relatively fre-quently in the Lupoaia macro£ora (Ticleanu andDiaconita, 1997). Many herbs and shrubs mayhave been present in such environments, e.g. As-teraceae p.p., Brassicaceae, Polygonaceae, Poa-ceae p.p., Buxus, etc. This vegetation may alsohave included some Mediterranean elements(Quercus ilex type, Phillyrea, Periploca, Pistacia,Olea, etc.), as observed in the present-day Colchidvegetation (southern edge of the Caucasus) (Denk

et al., 2002). Finally, the vegetation group con-tained a variety of megathermic elements atnon-signi¢cant percentages (Euphorbiaceae in-cluding Amanoa, Mimosaceae including Entada,Meliaceae, Sapindaceae, Loranthaceae, Areca-ceae, Sapotaceae, Tiliaceae).

(4) At higher altitudes, the composition of themixed forest would progressively change, with anincreasing presence of gymnosperms. First Ca-thaya and Sciadopitys would have been present,replaced subsequently by Cedrus and Tsuga, and¢nally by Abies and Picea. Traditionally, North-European palaeobotanists consider that thesetrees inhabited lowlands (in association with ther-mophilous elements) during the Early Pliocene, asdocumented by cones and seeds in the LowerRhine £uvial plain (Mai, 1995). Today, Abiesand Picea (still living in northern Europe) wouldbe present above 500 m altitude in this area, ac-cording to the latitudinal^altitudinal gradient con-trolling the modern distribution of trees in Europe(Ozenda, 1975, 1989). Therefore, during the EarlyPliocene, these conifers may have been repre-

Fig. 5. Principle component analysis applied to some key taxa and some ecological groups recorded in the Lupoaia pollen analy-sis. Axis 1 does not show any clear information. Axis 2 shows two groups which can be interpreted as a selection of the swamp-marsh elements with regard to the other groups. Taxa and groups which have been used in the calculations are those representa-tive of humid (climatic or edaphic) conditions. Pollen groups are those used in the synthetic pollen diagram (Fig. 4), despite thefollowing exceptions in order to test the relative contribution of some elements: (1) swamp trees include Taxodiaceae (Sciadopitysexcluded) and Nyssa ; (2) the mesothermic elements have been subdivided into two groups according to their local humidity re-quirements (this includes a riparian association containing Alnus, Salix, Populus, Liquidambar and Parrotia). Within the herbs,the two most abundant elements, Poaceae and Cyperaceae, have been tested separately. Megathermic and the other mega-meso-thermic elements have not been considered because of their low frequencies.

PALBO 2442 6-8-02


sented by somewhat ‘thermophilous’ species.However, the presence of Abies and Picea mightindicate cooler conditions in the Middle Pliocene(late Reuverian), when the increase in pollen per-centages of these two taxa occurs together with astrong decrease in thermophilous elements (Zag-wijn, 1960). At low European latitudes, EarlyPliocene £oristic conditions were rather di¡erentfrom northern Europe (Zagwijn and Suc, 1984).Indeed, no cones or seeds of Abies and Picea havebeen found in lowland or coastal lignites (RioMaior in Portugal: Diniz, 1984; Arjuzanx insouthwestern France: Huard, 1966; Cessenon insoutheastern France: Roiron, 1992), and this alsoapplies to the Romanian lignites (Husnicioara,Lupoaia, etc. : Ticleanu and Diaconita, 1997). Insouthern Europe, only Zanclean pollen diagramsfrom areas located at the foot of high massifs(Pyrenees, French Massif Central, Alps, Apen-nines) show high percentages of Cathaya, Cedrus,Tsuga, Abies and Picea (Suc et al., 1999). Similarrecords characterise southwestern Romania (Dri-valiari et al., 1999; Popescu, present work andunpublished data). In addition, signi¢cant in-creases in Abies and Picea represent the earliestNorthern Hemisphere glacials in pollen diagramsfrom the Apennines (Bertini and Roiron, 1997;Pontini and Bertini, 2000); this is also supportedby the Picea macrofossils found in the lignitequarry of Santa Barbara (Bertini and Roiron,1997). For these reasons, Abies and Picea are con-sidered to constitute a South-European montaneforest belt, even though its temperature range wassomewhat higher than its modern analogue in theEuropean massifs. Cedrus is considered as grow-ing in an intermediate altitudinal belt. As previ-ously discussed by Suc (1981), at Cessenon (footof the south French Massif Central) a lignite hasbeen found which is rich (20^40%) in Cedrus pol-len, but lacks any macrofossils belonging to thisgenus. Similarly, Cathaya, Keteleeria and Tsugacan be regarded as mid-altitude conifers.

5. Repetitive vegetation changes

The synthetic pollen diagram suggests repetitivechanges (Fig. 4A) between (1) a group consisting

of the thermophilous taxa (i.e. megathermic ele-ments (such as Mimosaceae, Meliaceae, Amanoa,Sapindaceae, Loranthaceae, etc.) and mega-meso-thermic elements such as Taxodiaceae (Sciado-pitys excluded), Juglans cf. cathayensis, Engel-hardia, Myrica, Nyssa cf. sinensis, Distylium,Parthenocissus cf. henryana, Parrotiopsis cf.jacquemontiana, Microtropis fallax, Cyrillaceae^Clethraceae, Leea, etc. plus Cupressaceae), and(2) a group containing the altitudinal trees (Tsuga,Cedrus, Cathaya, Abies and Picea). Relative fre-quencies of these groups are shown in Fig. 4B forcomparison with changes in lithology.

On the whole, maxima of thermophilous treescorrespond to lignite layers, whereas peaks in al-titudinal trees occur within clays. A straightfor-ward explanation of the cause of these alterna-tions might be based on di¡erences in pollentransport. Local vegetation may be well-repre-sented in lignite layers, whereas distant vegetationwould be better represented in clay layers as aresult of long-distance transport of terrigenousmaterial, including pollen, from altitudinal vege-tation belts. In this scenario, lignite depositionwould have been unpredictable, caused bychanges in the alluvial plain morphology resultingfrom shifts in the course of the river channel.There are, however, several arguments that con-tradict such an interpretation:

(1) The continuous lignite layers across the en-tire basin suggest that there were no local e¡ectsof changes in the river channel (Ticleanu and An-dreescu, 1988).

(2) Some thermophilous tree maxima occurwithin clay layers (below lignite VI, below ligniteVII, below lignite VIII and above lignite IX);some altitudinal tree maxima occur within lignitelayers (base of lignite VI, upper lignite VII, ligniteIX).

(3) Today, the pollen frequency of montanetrees (Abies and Picea) is not high in prodeltaicclayey areas, as seen in the northwestern Mediter-ranean region (Fauquette et al., 1999b).

(4) Mesophilous elements (deciduous Quercus,Carya, Pterocarya, Liquidambar, Carpinus, Parro-tia, etc.), which inhabited areas relatively distantfrom the swamps, show, on the whole, a constantabundance (ranging from 20 to 40% of the pollen

PALBO 2442 6-8-02


sum) independent of the clay^lignite alterna-tions.

(5) Similar variations in pollen spectra betweenthermophilous elements and altitudinal trees havebeen recorded in homogenous clay from EarlyPliocene sections in the Po Valley (Bertini, 1994).

(6) Close to Lupoaia, the Hinova section (low-ermost Pliocene according to nannoplankton evi-dence; Marunteanu, personal information), en-tirely constituted by clays, shows very lowpercentages of altitudinal trees (Popescu, unpub-lished).

(7) Pollen records between lignites IV and Vand above lignite VIII are characterised by rela-tively low frequencies of altitudinal trees, irrespec-tive of lithology (lignites or clays) ; on the con-trary, frequencies of thermophilous trees arehigher in the same intervals. In addition, pollenrecords between lignites IV and V include somemegathermic elements. This suggests that thethree main climatic phases of the Early Pliocenecould be represented in the Lupoaia section(Brunssumian A or phase P Ia in the Lupoaiaborehole F6 up to lignite V; Brunssumian B orphase P Ib between lignites V and VIII in theLupoaia quarry; Brunssumian C or phase P Icabove lignite VIII in the Lupoaia quarry; Sucand Zagwijn, 1983). The two warm Zanclean

phases (P Ia and P Ic) would be characterisedby a larger development of swamps in the DacicBasin while the Zanclean cooling phase (P Ib)would be characterised by reduced swamps anda moderate lowering of altitudinal vegetation beltsin the area. This is consistent with the larger ex-tension in space and in thickness of lignites V andVIII (Jipa, personal information).

On the basis of these arguments, a second andmore probable interpretation can be made of thelong-term thermophilous^altitudinal tree alterna-tions recorded at Lupoaia. These could well indi-cate successive warm^cool £uctuations corre-sponding to (1) the development of swampenvironments (warmer phases), (2) the descent tolower altitudes of the coniferous forests, with acorresponding increase in their pollen representa-tion in the delta sediments (cooler phases). A cor-relation analysis has been performed on percen-tages of thermophilous trees versus those ofaltitudinal trees for samples 1^110, i.e. in thepart of the Lupoaia section where clay^lignite al-ternations are very contrasted; the linear regres-sion is negative (Fig. 6), and supports this inter-pretation.

Such regular repetitions need to be juxtaposedwith £uctuations in eccentricity in order to testthe hypothesis of Van Vugt et al. (2001) who sug-gested a periodicity of 100 kyr for the lignitelayers. Fig. 7 focusses on the period consideredby Van Vugt et al. (2001) to correspond to theLupoaia section deposition. This period is charac-terised by great contrasts between eccentricitymaxima and minima. The global polarity timescale has been chronologically calibrated on thebasis of the eccentricity process. In turn, this timescale has been used to establish a chronology forthe N

18O curve, taken from a section in whichmagnetic reversals have been distinctly identi¢ed(Shackleton et al., 1995). The pollen curves maytherefore be related to the calibrated time scaleusing the palaeomagnetic reversals of the Lupoaiasection. This gives two independent lines of evi-dence to allow testing of the hypothesis of eccen-tricity forcing of lignite formation.

In general, the N18O curve indicates a relative

cooling phase corresponding to the C3n.2nChron. However, the pollen record indicates a

Fig. 6. Linear regression between percentages of thermophi-lous trees and altitude trees using the reduced major axismethod. Regression line is: % thermophilous trees =(32.228Ualtitude trees)+71.476 (slope: 32.228Y 0.522; inter-cept: 71.476Y 145.8). Unbiased correlation coe⁄cients are:Pearson’s r=30.199 (*P=0.0375); Spearman’s r=30.204(*P=0.033).

PALBO 2442 6-8-02


Fig. 7. Curves of thermophilous elements versus altitudinal trees (see Fig. 4B) are compared to eccentricity curve (Laskar, 1990) and to a reference N18O curve

(Shackleton et al., 1995) within the chronological frame proposed by Radan and Radan (1998) and Van Vugt et al. (2001). Some discrepancies between pollencurves and N

18O and/or eccentricity curves are indicated close to magnetic reversals for comparisons within similar accurate chronological intervals.

PALBO

24426-8-02

S.-M

.Popescu

/Review

ofPalaeobotany

andPalynology

120(2001)

181^202193

warm climate (high percentages of thermophiloustrees) during this normal polarity event. The in-terval between Chrons C3n.2n and C3n.1n corre-sponds to a warmer period according to the N

18Orecord; this is also supported by changes in eccen-tricity. However, this is in harmony with pollendata from the reverse interval in the lower part ofthe Lupoaia section, which are indicative of cool-er conditions (decrease in thermophilous ele-ments). The same line of reasoning cannot be ap-plied to the upper Lupoaia section, since thedi¡erences between the pollen record and N

18Ocurve are not so well marked for the time intervalbetween the early C3n.1n Chron and the earlyC2An.3n Chron. A more detailed examinationof the magnetic reversals shows some major dis-crepancies (points 1, 3, 4 below) between the pol-len record and eccentricity and/or N

18O curves,and one (point 2) of minor importance:

(1) The reversal at the end of the C3n.2n Chronoccurs just before the end of a prolonged maxi-mum in eccentricity and should therefore corre-spond to a general cooling period (see also theN18O reference curve). This is in contradiction tothe high abundance of thermophilous trees in theLupoaia pollen record, which exists for some timeprior to this period.

(2) The reversal at the onset of the C3n.1nChron corresponds to a transition from a maxi-mum eccentricity to a moderate minimum eccen-tricity, i.e. from a cooler to a warmer period. Thisis clearly seen in the N

18O reference curve, butappears to be in contradiction with high thermo-philous tree percentages in the Lupoaia pollenrecord. However, the imprecise de¢nition of theposition of this magnetic reversal reduces thestrength of this argument.

(3) Following this event, there is a reduction ineccentricity and the N18O curve indicates a warmerclimate, in contradiction with the relatively largepercentages of altitudinal trees in the Lupoaiapollen record.

(4) The upper reversal of C3n.1n Chron corre-sponds to the transition from a maximum eccen-tricity to a relative minimum eccentricity, i.e. fromcooler to relatively warmer climatic conditions(see also the N

18O reference curve). During thesame period, however, the Lupoaia pollen records

shows an evolution from warmer (high percen-tages in thermophilous trees) to cooler (increasein altitudinal trees) conditions.

Due to these incompatibilities, an attempt hasbeen made to correlate the Lupoaia section with asomewhat earlier period than proposed by VanVugt et al. (2001). In this concept, tested using asimilar comparative approach, the two normalpolarity episodes have been assigned to C3n.3nand C3n.2n Chrons, respectively, as indicated onFig. 8. The period is dominated by a long max-imum in eccentricity, interrupted by some briefrelative minima. This chronology gives an almostcomplete correspondence between the Lupoaiapollen curves and the global climatic referencecurves (eccentricity and N

18O). This is particularlyobvious at the magnetic reversals :

(1) The reversal at the end of the C3n.3n Chroncorresponds to a minimum eccentricity, indicatinga warm period that is supported by both N

18Ovalues and by large percentages of thermophiloustrees in the Lupoaia pollen record.

(2) The reversal at the onset of the C3n.2nChron corresponds to a short minimum eccentric-ity within a long maximum phase, i.e. to warmerconditions within a long cooling period. This is ingood agreement with the N18O curve and the max-imum of thermophilous trees between two maxi-ma of altitudinal trees in the Lupoaia pollen rec-ord.

(3) The reversal at the end of the C3n.2n Chronmatches the transition from minimum to maxi-mum eccentricity at the end of the long periodof maximum eccentricity. This is consistent withthe warmer phase expressed both by the N

18Oreference curve and thermophilous trees in theLupoaia pollen record.

Similar inferences may be made for all the ma-jor and minor concordances shown in Fig. 8.

It is therefore proposed that the previously es-timated correlation of the Lupoaia section (Ra-dan and Radan, 1998; Van Vugt et al., 2001)has to be changed to correspond with the normalChrons C3n.3n and C3n.2n. In this case, the evo-lution of the regional climate as re£ected by thepollen record is fully consistent with global cli-matic evolution (insolation and N

18O record).Such a chronological assignment is supported by

PALBO 2442 6-8-02


Fig. 8. Curves of thermophilous elements versus altitudinal trees (see Fig. 4B) are compared with the eccentricity curve (Laskar, 1990) and a reference N18O curve

(Shackleton et al., 1995) using a chronological frame, which is one chron older than that proposed by Radan and Radan (1998) and Van Vugt et al. (2001). Someconcordances between pollen curves and N

18O and eccentricity curves are indicated close to magnetic reversals for comparisons within similar accurate chronologicalintervals.

PALBO

24426-8-02

S.-M

.Popescu

/Review

ofPalaeobotany

andPalynology

120(2001)

181^202195

the mammal fauna, especially by the presence of aprimitive Mimomys in lignite VIII. In addition,the proposed age of the Lupoaia section wouldbe fully consistent with the climatic subdivisionof the Early Pliocene in Europe (Zagwijn, 1960;Suc, 1982; Suc and Zagwijn, 1983), in which acooling period (Brunssumian B=phase P Ib) isobserved between two warm periods (Fig. 9).

The Lupoaia section belongs almost completelyto this cooling period, which has also been ob-served in the Ticleni borehole. This is further sup-ported by the regional lignite stratigraphy, coolingbeing centred around lignite VII (Drivaliari et al.,1999) (Fig. 9).

The Van Vugt et al. (2001) hypothesis is con-¢rmed: lignite layers seem to occur every 100 kyr.

Fig. 9. Place of the Lupoaia pollen diagram in the frame of the regional (Ticleni: Drivaliari et al., 1999) and European climato-stratigraphies (Susteren in The Netherlands: Zagwijn, 1960; Garraf 1 in the northwestern Mediterranean region: Suc and Cra-vatte, 1982) and the reference global N18O curve (Shackleton et al., 1995). Climatostratigraphic relationships are based on the re-spective curves of thermophilous trees focussing on the warm Early Pliocene phase (Brunssumian in The Netherlands, pollenphase P I in the Mediterranean). The Lupoaia pollen diagram covers the upper Brunssumian A (pollen zone P Ia in the Mediter-ranean) to the lower Brunssumian C (pollen zone P Ic in the Mediterranean). Position of lignites IV, VII and IX^XI complex isindicated in support of climatostratigraphic relationships between the Ticleni and the Lupoaia sections.

Fig. 10. Cyclic responses in the Lupoaia pollen record of thermophilous and altitudinal trees to eccentricity. Chronology of pa-laeomagnetic reversals after Lourens et al. (1996). Shaded areas represent warm parts of the climatic cycles inferred from compar-ison of thermophilous tree maxima with eccentricity minima. Lithology: see Fig. 2.

PALBO 2442 6-8-02


PALBO 2442 6-8-02


They are linked to minimum eccentricity. A sim-ple estimate, based on the time separating the twomost distant magnetic reversals obviously re-corded in the Lupoaia section (4.799 (end ofC3n.3n Chron)^4.493 Ma (end of C3n.2n Chron)according to Lourens et al., 1996), indicates aperiod of 306 kyr. This period includes three max-ima of altitudinal trees and three relative maximaof eccentricity (100-kyr period) (Fig. 10). A spec-tral analysis of thermophilous elements and alti-tudinal trees has been performed with respect tothickness of the section (Analyseries Program:Paillard et al., 1996). Both groups show a strongspectral peak which has been calculated as locatedat a thickness of 20 m (Fig. 11). This corresponds

to a time period of approximately 87 kyr, and canbe considered consistent with the period of eccen-tricity, when the changes in sedimentation ratethat probably di¡erentiated lignites from claysare taken into account.

Pollen concentration can provide informationon preservation and/or changes in sedimentationrate (Suc, 1984b); low values may be associatedwith higher sedimentation rate (pollen grains di-luted within a lot of terrigenous particles), andhigh values with a low sedimentation rate. Fig.12 shows variations in pollen concentration alongthe Lupoaia section. The highest values (s 10 000pollen grains/g of sediment) mostly correspond tolignite layers, with the exception of some samplesfrom borehole F6. Examination of the respectivethickness of clays and lignites within each eccen-tricity cycle (sands are excluded because theyprobably represent a short time-interval), showsthat lignites represent less than half a cycle, whereone cycle equals the clay layer and the overlyinglignite (Fig. 10): lignite V= 25% of the cycle; lig-nite VI= 40% of the cycle; lignite VII = 25% ofthe cycle; lignite VIII = 35% of the cycle. Forthree of these lignites (V, VI, VII), the relatedminimum eccentricity is considerably shorter intime than the maximum part of the eccentricitycurve corresponding to the underlying clays.However, lignite VIII seems thinner as may beexpected on the basis of the eccentricity curve,which shows a minimum of longer duration. Incontrast, the thickness of lignites IX plus X^XIII(s 60% of the cycle) seems to be in agreementwith the relatively long duration of the corre-sponding minimum in the eccentricity curve. Thepollen concentration appears to vary independent-ly of the variation in thickness within the cycles(high pollen concentration for lignites VI, VII andX^XIII; low pollen concentration for lignites Vand VIII). As a consequence, pollen concentrationseems to be more linked to preservation and pos-sibly to xylite abundance in some lignites. Thismay explain the scarcity, and occasional absenceof pollen grains in some samples.

The eccentricity-forced pollen cycles (thermo-philous plants versus altitudinal trees) and lithol-ogy cycles (clay^lignite alternations) suggest thatthe inferred ages throughout the studied section

Fig. 11. Spectral analysis of thermophilous elements and alti-tude trees. E, peak considered as representative of eccentric-ity period; O, peak considered as representative of obliquityperiod. The most signi¢cant peak E corresponds to 0.0005 asfrequency. The correlative period is inverse to frequency (ex-pressed in cm), i.e. 2000 cm.

PALBO 2442 6-8-02


Fig. 12. Pollen concentration along the Lupoaia section. Lithology: see Fig. 2.

PALBO 2442 6-8-02


correspond well with respect to the eccentricitychronology. As a consequence, it is proposedthat (1) the undetermined polarity zone in thebasal part of the section should belong to thereversed polarity interval prior to C3n.3n Chron(the reversal should occur within the clays locatedabove lignite IV), (2) the reversal at the end ofChron C3n.3n should occur within the undeter-mined polarity zone, more precisely at the topof lignite V, and (3) the undetermined polarityzone located at the top of the Lupoaia sectionshould immediately precede the following normalinterval, i.e. the C3n.1n Chron (Fig. 10). The lat-ter inference is based on both the eccentricitychronology and the close relationships betweenthe N

18O reference curve and the thermophiloustree curve in the Lupoaia pollen record (Fig. 9).Taking this information into account, the studiedLupoaia section would represent a time span ofabout 600 kyr, i.e. from about 4.90 to about 4.30Ma. The three climatic phases of the Early Plio-cene in the Lupoaia pollen record (see above:warm phase ending with lignite V, cooler phasebetween lignites V and VIII, warmer phase start-ing with lignite VIII; Fig. 10) may be the expres-sion of the 400-kyr cycles of eccentricity over-printing the 100-kyr cycles (see Figs. 8 and 10).

6. Conclusions

High-resolution pollen analyses of the Lupoaiasection (borehole F6 and quarry) have (1) allowedthe reconstruction of the Early Pliocene vegeta-tion of southwestern Romania, and (2) providedevidence of repetitive changes in vegetation. Theresults show two opposed vegetation belts : ther-mophilous trees, which were located in low alti-tudes of the Dacic Basin, and altitudinal treeswhich probably grew in the Carpathians. Duringcooling periods, this second group migrated tolower altitudes, and their pollen grains partlymasked the local production in the alluvial plain.In this region, lignite deposition corresponds towarm phases under continuously humid condi-tions.

These vegetation changes and the clay^lignitealternations are forced by eccentricity. A detailed

comparison between pollen data and global cli-matic £uctuations (provided by eccentricity andN18O curves) has led to a change in the chrono-logical assignment of the section, previously basedon a preliminary interpretation of palaeomagneticmeasurements (Radan and Radan, 1998; VanVugt et al., 2001). The revised chronological as-signment is in accordance with the mammal fau-na, whereas the regional climatic £uctuations areconsistent with the global climatic evolution. TheLupoaia section covers a period between approx-imately 4.90 and 4.30 Ma and represents a timespan of about 600 kyr. The section completelyincludes the C3n.3n and C3n.2n Chrons. In addi-tion, it is proposed that the section ends at theonset of the C3n.1n Chron.

Climatically, the Lupoaia section covers:(1) The end of the Early Pliocene warm phase

(Brunssumian A in northern Europe and P Iaphase in the northwestern Mediterranean region;Suc and Zagwijn, 1983).

(2) The entire Early Pliocene cooling (Bruns-sumian B in Northern Europe and P Ib phase inthe Northwestern Mediterranean region; Suc andZagwijn, 1983).

(3) The beginning of the late Early Pliocenewarm phase (Brunssumian C in Northern Europeand P Ic phase in the Northwestern Mediterra-nean region; Suc and Zagwijn, 1983).

The main Early Pliocene climatic subdivisionsare now documented for the east of Europe butappear more complex. Other areas (for example atthe base of mountains) may provide similarly de-tailed records when studied with high-resolutionpollen investigation.

Acknowledgements

This Ph.D. work was granted by the FrenchGovernment through its Embassy at Bucharest.This study was supported by the Programme ‘En-vironnement, Vie et Socie¤te¤s’. J.-P. Suc and P.Bernier have supervised this work. The manu-script has been improved thanks to the commentsof the two referees, Dr. S. Bottema and Dr. V.Mosbrugger. My friends, the Lupoaia quarry en-gineers, provided facilities for sampling. S.C. Ra-

PALBO 2442 6-8-02


dan gave samples from borehole F6. G. Clauzon,D. Jipa and M. Marunteanu supplied informationon palaeogeography and stratigraphy in theTurnu Severin area. I received help from M. Gon-zales for performing samples, from G. Escargueland F. Giraud for statistic treatments. I got a lotof information from P. Mein, C. Gue¤rin, J. Agus-ti, A.J. Van der Meulen, M. Fortelius and M.Erbajeva on the mammal fauna. P. Bernier, S.Legendre and J.-P. Suc discussed the manuscript,the English edition of which has been made byDr. Simon Brewer.

References

Aguilar, J.-P., Legendre, S., Michaux, J., Montuire, S., 1999.Pliocene mammals and climatic reconstruction in the West-ern Mediterranean area. In: Wrenn, J., Suc, J.-P., Leroy,S.A.G. (Eds.), The Pliocene: Time of Change. Am. Assoc.Stratigr. Palynol. Found., pp. 109^120.

Aguilar, J.-P., Michaux, J., 1984. Le gisement a' micromammi-fe'res du Mont-He¤le'ne (Pyre¤ne¤es-Orientales): apports a' laconnaissance de l’histoire des faunes et des environnementscontinentaux. Implications stratigraphiques pour le Plioce'nedu sud de la France. Pale¤obiol. Cont. 14, 19^31.

Apostol, L., Enache, C., 1979. Etude de l’espe'ce Dicerorhinusmegarhinus (de Christol) du bassin carbonife're de Motru.Trav. Mus. Hist. Nat. Grigore Antipa 20, 533^540.

Bertini, A., 1994. Messinian-Zanclean vegetation and climatein North-Central Italy. Hist. Biol. 9, 3^10.

Bertini, A., Roiron, P., 1997. Evolution de la ve¤ge¤tation et duclimat pendant le Plioce'ne moyen, en Italie centrale: apportde la palynologie et de la macro£ore a' l’e¤tude du bassin duValdarno supe¤rieur (coupe de Santa Barbara). C. R. Acad.Sci. Paris 324, 763^771.

Cour, P., 1974. Nouvelles techniques de de¤tection des £ux etdes retombe¤es polliniques: e¤tude de la se¤dimentation despollens et des spores a' la surface du sol. Pollen Spores 16,103^141.

Denk, T., Frotzler, N., Davitashvili, N., 2002. Vegetationalpatterns and distribution of relict taxa in humid temperateforests and wetlands of Georgia (Transcaucasia). Biol. J.Linnean Soc., in press.

Diniz, F., 1984. Apports de la palynologie a' la connaissancedu Plioce'ne portugais. Rio Maior: un bassin de re¤fe¤rencepour l’histoire de la £ore, de la ve¤ge¤tation et du climat de lafacRade atlantique de l’Europe me¤ridionale. Thesis, Univer-site¤ Montpellier 2, p. 230.

Drivaliari, A., 1993. Images polliniques et pale¤oenvironnementau Ne¤oge'ne supe¤rieur en Me¤diterrane¤e orientale. Aspectsclimatiques et pale¤oge¤ographiques d’un transect latitudinal(de la Roumanie au delta du Nil). Thesis, Universite¤ Mont-pellier 2, p. 332.

Drivaliari, A., Ticleanu, N., Marinescu, F., Marunteanu, M.,Suc, J.-P., 1999. A Pliocene climatic record at Ticleni(Southwestern Romania). In: Wrenn, J.H., Suc, J.-P., Leroy,S.A.G. (Eds.), The Pliocene: Time of Change. Am. Assoc.Stratigr. Palynol. Found., pp. 103^108.

Fauquette, S., Clauzon, G., Suc, J.-P., Zheng, Z., 1999b. Anew approach for paleoaltitude estimates based on pollenrecords: example of the Mercantour Massif (southeasternFrance) at the earliest Pliocene. Earth Planet. Sci. Lett.170, 35^47.

Fauquette, S., Guiot, J., Suc, J.-P., 1999a. A method for cli-matic reconstruction of the Mediterranean Pliocene usingpollen data. Palaeogeogr. Palaeoclimatol. Palaeoecol. 144,183^201.

George, J.C., 1972. Everglades Wildguide. Natural History Se-ries, Nat. Park Service, US Department of the Interior.

Gue¤rin, C., 1980. Les Rhinoceros (Mammalia, Perissodactyla)du Mioce'ne terminal au Ple¤istoce'ne supe¤rieur en Europeoccidentale. Comparaison avec les espe'ces actuelles. Doc.Sci. Terre Universite¤, 79, Claude-Bernard Lyon 1, pp. 1^784.

Hilgen, F.J., 1990. Sedimentary cycles and an astronomicallycontrolled, oscillatory system of climatic changes during theLate Cenozoic in the Mediterranean. Pale¤obiol. Cont. 17,25^33.

Hilgen, F.J., 1991a. Astronomical calibration of Gauss toMatayama sapropels in the Mediterranean and implicationfor the Geomagnetic Polarity Time Scale. Earth Planet. Sci.Lett. 104, 226^244.

Hilgen, F.J., 1991b. Extension of the astronomically calibrated(polarity) time scale to the Miocene/Pliocene boundary.Earth Planet. Sci. Lett. 107, 349^368.

Huard, J., 1966. Etude anatomique des bois de conife'res descouches a' lignite ne¤oge'nes des Landes. Me¤m. Soc. Ge¤ol. Fr.105, 7^85.

Kloosterboer-van Hoeve, M., 2000. Cyclic changes in the lateNeogene vegetation of northern Greece. LPP Contr. Ser. 12,1^131.

Koufos, G.D., Pavlides, S.B., 1988. Correlation between thecontinental deposits of the lower Axios valley and Ptolemaisbasin. Bull. Geol. Soc. Greece 20, 9^19.

Laskar, J., 1990. The chaotic motion of the solar system: anumerical estimate of the size of the chaotic zones. Ikarus88, 266^291.

Lourens, L.J., Antonarakou, A., Hilgen, F.J., van Hoof,A.A.M., Vergnaud Grazzini, C., Zachariasse, W.J., 1996.Evaluation of the Plio-Pleistocene astronomical time scale.Paleoceanography 11, 391^413.

Mai, D.H., 1995. In: Fischer, G. (Ed.), Tertia«re Vegetations-geschichte Europas. Methoden und Ergebnisse. Jean, Stutt-gart, New York, p. 691.

Menke, B., 1975. Vegetationsgeschichte und Florenstratigra-phie Nordwest-Deutschlands im Plioza«n und Fru«hquarta«r.Mit einem Beitrag zur Biostratigraphie des Weichselfru«hgla-zials. Geol. Jb. A 26, 3^151.

Ozenda, P., 1975. Sur les e¤tages de ve¤ge¤tation dans les mon-tagnes du bassin me¤diterrane¤en. Doc. Cart. Ecol. 16, 1^32.

Ozenda, P., 1989. Le de¤placement vertical des e¤tages de ve¤ge¤-

PALBO 2442 6-8-02


tation en fonction de la latitude: un mode'le simple et seslimites. Bull. Soc. Ge¤ol. France 5, 535^540.

Paillard, D., Labeyrie, L., Yiou, P., 1996. Macintosh Programperforms time-series analysis. Eos Trans. AGU, 379.

Petrescu, I., Nica, T., Filipescu, S., Barbu, O., Chira, C., Av-ram, R., Valaczkai, T., 1989. Paleoclimatical signi¢cance ofthe palynological approach to the Pliocene deposits of Lu-poaia (Gorj county). Stud. Univ. Babeu¤-Bo¤lyai Geol.-Geogr. 34, 75^81.

Pontini, M.R., Bertini, A., 2000. Late Pliocene vegetation andclimate in Central Italy: high-resolution pollen analysesfrom the Fosso Bianco succession (Tiberino Basin). Geobios33, 519^526.

Radan, S.C., Radan, M., 1998. Study of the geomagnetic ¢eldstructure in the Tertiary in the context of magnetostrati-graphic scale elaboration. I: The Pliocene. An. Inst. Geol.Rom. 70, 215^231.

Radulescu, C., Samson, P.-M., Sen, S., Stiuca, E., Horoi, V.,1997. Les micrommmife'res plioce'nes de Dranic (bassin Da-cique, Roumanie). In: Aguilar, J.-P., Michaux, J., Legendre,S. (Eds.), Actes du Congre's BiochroM’97. Me¤m. Trav.E.P.H.E., Inst. Montpellier, 21, pp. 635^647.

Rich, F.J., 1985. Palynology and Paleoecology of a LigniticPeat from Trail Ridge, Florida. Information Circular 100.Florida Geol. Survey, pp. 1^15.

Roberts, H.H., 1986. Selected depositional environments of theMississippi River deltaic plain. Geol. Soc. of America Cen-tennial Field Guide, Southeastern Sect., pp. 435^440.

Roiron, P., 1992. Flores, ve¤ge¤tations et climats du Ne¤oge'neme¤diterrane¤en: apports de macro£ores du sud de la Franceet du nord-est de l’Espagne. Thesis, Universite¤ Montpellier 2.

De Serres, M., 1819. Observations sur divers fossiles de qua-drupe'des vivipares nouvellement de¤couverts dans le sol desenvirons de Montpellier. J. Phys. 88, 382, 394, 405, 417.

Shackleton, N.J., Hall, M.A., Pate, D., 1995. Pliocene stableisotope stratigraphy of Site 846. Proc. Ocean Drill. Progr.138, 337^355.

Suc, J.-P., 1981. La ve¤ge¤tation et le climat du Languedoc (sudde la France) au Plioce'ne moyen d’apre's la Palynologie.Pale¤obiol. Cont. 12, 7^26.

Suc, J.-P., 1982. Palynostratigraphie et pale¤oclimatologie duPlioce'ne et du Ple¤istoce'ne infe¤rieur en Me¤diterrane¤e nord-occidentale. C. R. Acad. Sci. Paris 294, 1003^1008.

Suc, J.-P., 1984a. Origin and evolution of the Mediterraneanvegetation and climate in Europe. Nature 307, 429^432.

Suc, J.-P., 1984b. La signi¢cation pale¤oe¤cologique des restesve¤ge¤taux. Bull. Centres rech. Explor.-Prod. Elf-Aquitaine 8,97^104.

Suc, J.-P., Cravatte, J., 1982. Etude palynologique du Plioce'nede Catalogne (nord-est de l’Espagne). Pale¤obiol. Cont. 13,1^31.

Suc, J.-P., Diniz, F., Leroy, S., Poumot, C., Bertini, A., Du-pont, L., Clet, M., Bessais, E., Zheng, Z., Fauquette, S.,Ferrier, J., 1995. Zanclean (VBrunssumian) to early Piacen-zian (Vearly-middle Reuverian) climate from 4‡ to 54‡north latitude (West Africa, West Europe and West Medi-terranean areas). Meded. Rijks Geol. Diesnt. 52, 43^56.

Suc, J.-P., Fauquette, S., Bessedik, M., Bertini, A., Zheng, Z.,Clauzon, G., Suballyova, D., Diniz, F., Que¤zel, P., Feddi,N., Clet, M., Bessais, E., Bachiri Taou¢q, N., Me¤on, H.,Combourieu-Nebout, N., 1999. Neogene vegetation changesin West European and West circum-Mediterranean areas.In: Agusti, J., Rook, L., Andrews, P. (Eds.), HominoidEvolution and Climatic Change in Europe, vol. 1: The Evo-lution of Neogene Terrestrial Ecosystems in Europe. Cam-bridge University Press, Cambridge, pp. 378^388.

Suc, J.-P., Zagwijn, W.H., 1983. Plio-Pleistocene correlationsbetween the northwestern Mediterranean region and north-western Europe according to recent biostratigraphic and pa-leoclimatic data. Boreas 12, 153^166.

Ticleanu, N., 1992. Main coal-generating paleophytocoenosesin the Pliocene of Oltenia. Rom. J. Paleontol. 75, 75^80.

Ticleanu, N., Andreescu, I., 1988. Considerations on the devel-opment of Pliocene coaly complexes in the Jiu-Motru sector(Oltenia). D. S. Inst. Geol. Geo¢z. 72^73 (2), 227^244.

Ticleanu, N., Buliga, S., 1992. Paleophytocoenotic researchesin Pliocene deposits from the Plostina zone (Gorj district).Rom. J. Paleontol. 75, 81^85.

Ticleanu, N., Diaconita, D., 1997. The main coal facies andlithotypes of the Pliocene coal basin, Oltenia, Romania. In:Gayer, R., Pesek, J. (Eds.), European Coal Geology andTechnology. Geol. Soc. Spec. Publ. 125, pp. 131^139.

Tiedemann, R., Sarnthein, M., Shackleton, N.J., 1994. Astro-nomical timescale for the Pliocene Atlantic N

18O and dust£ux records of ODP Site 659. Paleoceanography 9, 619^638.

Van der Zwaan, G.J., Gudjonsson, L., 1986. Middle Miocene-Pliocene stable isotope stratigraphy and paleoceanographyof the Mediterranean. Mar. Micropaleontol. 10, 71^90.

Van Vugt, N., Langereis, C.G., Hilgen, F.J., 2001. Orbitalforcing in Pliocene-Pleistocene Mediterranean lacustrine de-posits: Dominant expression of eccentricity versus preces-sion. Palaeogeogr. Palaeoclimatol. Palaeoecol. 172, 193^205.

Van Vugt, N., Steenbrink, J., Langereis, C.G., Hilgen, F.J.,Meulenkamp, J.E., 1998. Magnetostratigraphy-based astro-nomical tuning of the early Pliocene lacustrine sediments ofPtolemais (NW Greece) and bed-to-bed correlation with themarine record. Earth Planet. Sci. Lett. 164, 535^551.

Wang, C.W., 1961. The Forests of China with a Survey ofGrassland and Desert Vegetation, vol. 5. M. Moors CabotFoundation, p. 313.

Van de Weerd, A., 1978. Early Ruscinian rodents and lago-morphs (Mammalia) from the lignites near Ptolemais (Mac-edonia, Greece). Proc. K. Ned. Akad. Wet. B82, 127^170.

Zagwijn, W.H., 1960. Aspects of the Pliocene and early Pleis-tocene vegetation in The Netherlands. Meded. Geol. Sticht.C 3, 1^78.

Zagwijn, W.H., Suc, J.-P., 1984. Palynostratigraphie du Plio-Ple¤istoce'ne d’Europe et de Me¤diterrane¤e nord-occidentales:corre¤lations chronostratigraphiques, histoire de la ve¤ge¤tationet du climat. Pale¤obiol. Cont. 14, 475^483.

Zheng, Z., Cravatte, J., 1986. Etude palynologique du Plioce'nede la Co“te d’Azur (France) et du littoral ligure (Italie). Geo-bios 19, 815^823.

PALBO 2442 6-8-02


repetitive changes in early pliocene vegetation revealed ...€¦ · ent for north-european to...

Documents