Palynological evidence for Miocene climate change in the Forecarpathian Basin (Central Paratethys, NW Bulgaria)

Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/223919595

PalynologicalevidenceforMioceneclimatechangeintheForecarpathianBasin(CentralParatethys,NWBulgaria)

ArticleinPalaeogeographyPalaeoclimatologyPalaeoecology·February2002

DOI:10.1016/S0031-0182(01)00365-0

CITATIONS

145

READS

236

4authors,including:

Someoftheauthorsofthispublicationarealsoworkingontheserelatedprojects:

NECLIME-NeogeneClimateEvolutioninEurasiaViewproject

DimiterIvanov

BulgarianAcademyofSciences

66PUBLICATIONS895CITATIONS

SEEPROFILE

VolkerMosbrugger

SenckenbergResearchInstitute

158PUBLICATIONS5,123CITATIONS

SEEPROFILE

AllcontentfollowingthispagewasuploadedbyDimiterIvanovon11January2014.

Theuserhasrequestedenhancementofthedownloadedfile.Allin-textreferencesunderlinedinblue

arelinkedtopublicationsonResearchGate,lettingyouaccessandreadthemimmediately.

Palynological evidence for Miocene climate change in theForecarpathian Basin (Central Paratethys, NW Bulgaria)

D. Ivanov a, A.R. Ashraf b;*, V. Mosbrugger b, E. Palamarev a

a Bulgarian Academy of Sciences, Institute of Botany, Acad. G. Bonchev Str. 23, BG-1113 So¢a, Bulgariab Universita«t Tu«bingen, Institut fu«r Geologie und Pala«ontologie, Sigwartstr. 10, D-72076 Tu«bingen, Germany

Received 7 July 2000; accepted 29 August 2001

Abstract

We reconstruct quantitatively the Middle to Upper Miocene climate evolution in the southern ForecarpathianBasin (Central Paratethys area, Northwest Bulgaria) by applying the coexistence approach to 101 well-datedpalynofloras isolated from three cores. The climatic evolution is compared with changes in vegetation andpalaeogeography. The Middle Miocene was a period of a subtropical/warm^temperate humid climate with meanannual temperature (MAT) between 16 and 18‡C and mean annual precipitation (MAP) between 1100 and 1300 mm.Thereby, during the entire Middle Miocene a trend of slightly decreasing temperatures is observed and only smallclimate fluctuations occur which are presumably related to palaeogeographic reorganisations. The vegetation shows acorresponding trend with a decrease in abundance of palaeotropic and thermophilous elements. The Upper Miocene ischaracterised by more diverse climatic conditions, probably depending on palaeogeographic and global climatictransformations. The beginning of this period is marked by a slight cooling and a significant drying of the climate, withMAT 13.3^17‡C and MAP 652^759 mm. After that, fluctuations of all palaeoclimate parameters occur displayingcycles of humid/dryer and warmer/cooler conditions, which are again well reflected in the vegetation. Our studyprovides a first quantitative model of the Middle^Upper Miocene palaeoclimate evolution in Southeastern Europe andis characterised by a relatively high precision and resolution with respect to the climate data and stratigraphy. ß 2002Elsevier Science B.V. All rights reserved.

Keywords: Miocene; Palynology; Palaeoclimate; Vegetation; Bulgaria; Southeastern Europe

1. Introduction

The Forecarpathian Basin represents the east-ern part of the Central Paratethys (Fig. 1) and is akey region to understand the Neogene evolutionof the connection between the Central and East-

ern Paratethys area (Ro«gl, 1998). Apparently, theForecarpathian Basin also plays a major role inthe evolution and migration of Mediterraneanvegetation (Palamarev, 1989). The palaeogeogra-phy of the Forecarpathian Basin and its varia-tions during the Neogene are relatively wellknown (Kojumdgieva and Popov, 1986, 1989;Ro«gl, 1998). The Middle Miocene (Sarmatian) £o-ras of this area have also been studied on the baseof leaf imprints, seeds/fruits and dispersed cuticles(Hadgiev and Palamarev, 1962; Stefanov et al.,

0031-0182 / 02 / $ ^ see front matter ß 2002 Elsevier Science B.V. All rights reserved.PII: S 0 0 3 1 - 0 1 8 2 ( 0 1 ) 0 0 3 6 5 - 0

* Corresponding author. Fax: +49-7071-29-5727.E-mail address: [email protected]

(A.R. Ashraf).

PALAEO 2724 11-4-02

Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^37

www.elsevier.com/locate/palaeo

Fig. 1. (a) Palaeogeography of the Pannonian and Forecarpathian Basin during the Miocene (after Petkova, 1977). (b) Sketchmap showing the structural/palaeogeographic areas in Northwest Bulgaria during the Neogene as well as the locations of coresC-1, C-12 and C-37 (redrawn from Kojumdgieva and Popov, 1988). Legend: (1) areas outside the Forecarpathian basin (land);(2) marginal stable area; (3) Miocene longitudinal depression; (4) Lom depression; (5) boundary of the basin; (6) boundary ofthe Miocene longitudinal depression; (7) boundary of the Lom depression.

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^3720

1964; Palamarev and Uzunova, 1970, 1992; Pa-lamarev et al., 1975, 1978; Petkova, 1967; Petko-va and Kitanov, 1965; Palamarev, 1988, 1990,1993; Palamarev and Petkova, 1987; Uzunova,1995, 1996). More recently palynological studieshave been undertaken in the Neogene of theSouthern Forecarpathian Basin in Northwest Bul-garia (Ivanov, 1995, 1997). The studied palyno-morph assemblages originate from marine tobrackish sediments which are Middle to UpperMiocene in age as dated by foraminifera, molluscsand ostracods (Kojumdgieva et al., 1989).

In the present study we have re-analysed thepalynological data of Ivanov (1995, 1997) withthe help of the coexistence approach (Mosbrug-ger, 1995; Mosbrugger and Utescher, 1997) inorder to obtain quantitative data about the Neo-gene climate evolution in the Southern Forecarpa-thian Basin. In addition, we brie£y summarise thepresent knowledge of Neogene changes in palaeo-geography and vegetation in this area to betterunderstand the coupling between the dynamicsof climate, vegetation and palaeogeography.This will also help to distinguish global from re-gional climate change.

2. Study area, materials and methods

Our study area is located in NW Bulgaria andrepresents the southernmost part of the Forecar-pathian Basin (Fig. 1). The Neogene evolution ofthe basin is presented in detail by Kojumdgievaand Popov (1989), Kojumdgieva and Popov(1988) summarise the lithostratigraphy of thesediments and Kojumdgieva et al. (1989) providea correlation scheme for the biostratigraphic sub-divisions based on molluscs, foraminifera and os-tracods. This stratigraphic information is summa-rised in Table 1.

Our study of the climate and vegetation dynam-ics in the Southern Forecarpathian Basin is basedon three drillings, i.e. Drenovets C-1, Deleina C-12, and Makresh C-37, the locations of which areshown in Fig. 1b. All three cores are situated inthe so-called ‘Miocene longitudinal depression’which is characterised by marine, brackish andeven continental sediments. Thereby the Lower

to Middle Badenian deposits predominantly rep-resent normal marine environments as proven byforaminifera associations. During the Upper Ba-denian, the faunal composition is generally rela-tively poor and indicates variable salinity with anincreased fresh water in£ux. Based on molluscassociations the salinity of the Sarmatian deposi-tional environments is assumed to be mostlyaround 14^18x. Similarly, a brackish environ-ment predominates during the Maeotian andLower Pontian whereas the Upper Pontian andLower Pliocene are only represented by continen-tal deposits with a restricted geographic distribu-tion (for more details see Kojumdgieva and Po-pov, 1989).

The lithology of the cores is shown in Fig. 2and described in detail in Ivanov (1995). As indi-cated by foraminifera, molluscs and ostracods, thecores cover the stratigraphic range from the Low-er Badenian to Lower Pontian (Fig. 2, Table 1);major unconformities occur only at the base ofthe Volhynian (Lower Sarmatian) and of theMaeotian. The overall correlation of the coresrelies on these unconformities and on informationprovided by marine faunas as well as by palyno-morph assemblages (see Ivanov, 1995, 1997;Table 1, Fig. 2). A particular problem, however,is the detailed correlation of the relatively incom-plete core C-37. Whereas the entire Badenian isrepresented in core C-12 (although only the Mid-dle and Upper Badenian is documented by paly-no£oras; see Fig. 2), core C-37 covers only partsof the Upper Badenian. More speci¢cally, thislatter core does not reach down to the gypsumlayer which exists near the very base of the UpperBadenian in core C-12 (see Fig. 2). In addition,palynological evidence (Ivanov, 1995, 1997) aswell as the vegetation and climate evolution (seeSections 4 and 5) indicate that sedimentation inthe uppermost Upper Badenian stops earlier incore C-37 than in core C-12. This may be ex-plained by the fact that core C-37 is situated clos-er to the ancient shoreline.

The palyno£oras of the cores have been ¢guredand described in detail by Ivanov (1995, 1997)based on 39 samples from core Drenovets C-1,39 samples from core Deleina C-12 and 23 sam-ples from core Makresh C-37. The location of the

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^37 21

Table 1Stratigraphic chart and correlation of biostratigraphic subdivision of the Neogene in NW Bulgaria (slightly modi¢ed after Ko-jumdgieva et al., 1989)

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^3722

samples within the cores is shown in Fig. 2; sim-pli¢ed pollen diagrams are given in Figs. 3^5.Based on these 101 palyno£oras we have recon-structed the climate evolution using the so-calledcoexistence approach (Mosbrugger and Utescher,1997) This technique is straightforward and re-quires two steps. First, for all taxa of a givenfossil £ora the nearest living relatives and theirclimatic tolerances with respect to various climaticparameters are determined; then for the variousclimate parameters the interval is calculated with-in which all nearest living relatives of the fossil£ora can coexist. This coexistence interval is con-sidered to represent a reasonable estimator of thepast climate under which the fossil £ora lived. Thecoexistence approach can be used for all kinds of£oras (palyno£oras, leaves, fruits/seeds). For itsapplication a computer program is availablewhich extracts for a given fossil £ora the relevantinformation (nearest living relatives of fossil taxa,climatic tolerances of these nearest living rela-tives) from a database and then calculates thecoexistence intervals for various climate parame-ters (for details see Mosbrugger and Utescher,1997).

Of course, the coexistence approach implies anumber of uncertainties or sources of error: (1)the determination of fossil taxa may be incorrect ;(2) the assignment of a nearest living relative to afossil taxon may be incorrect ; (3) the descriptionof the climatic tolerances of the nearest living rel-atives may be wrong; (4) the climatic tolerance ofa nearest living relative may di¡er from the cli-matic tolerance of the corresponding fossil taxon.Mostly, these errors can be identi¢ed because theylead to inconsistencies or ‘outliers’ when the coex-istence interval is calculated (see Mosbrugger andUtescher, 1997 for details). Sometimes the appli-cation of the coexistence approach to a fossil £oracan also lead to two distinct coexistence intervals.This may result from one or several of the abovementioned errors or it may be caused by a mix-ture of distinct (allochthonous) £oras representingdi¡erent climate situations.

In our study the coexistence approach was ap-plied to all 101 palyno£oras of cores C-1, C-12,C-37 described in Ivanov (1995, 1997). Therebywe considered the following climate parameters:

^ MAT: mean annual temperature (‡C).^ TCM: mean temperature of the coldest

month (‡C).^ TWM: mean temperature of the warmest

month (‡C).^ MAP: mean annual precipitation (mm).As an additional climate proxy we determined

for all palyno£oras the relative proportion of pa-laeotropical (P) and arctotertiary (A) elements.According to classical de¢nitions (e.g. Mai,1995) the term ‘arctotertiary elements’ is usedfor plants which grew in the arctic area duringthe Paleogene under a temperate to warm^tem-perate climate and correspondingly occur todayin the temperate zone. In contrast, ‘palaeotropicelements’ are plants which have their present-daydistribution primarily in the palaeotropic area, i.e.in the tropical regions of Asia and Africa. Follow-ing the concept of other authors (e.g. Palamarevand Ivanov, 2001) we further subdivided the pa-laeotropical and arctotertiary elements into trop-ical (P1) and subtropical (P2) and into warm(A1)- and cool-temperate (A2) elements, respec-tively. This subdivision, however, did not provideadditional information and hence is not furtherdiscussed in the text.

3. Palaeogeographic evolution of the basin

The following summary of the palaeogeo-graphic evolution of the Forecarpathian Basis islargely based on data and results presented byKojumdgieva and Popov (1988, 1989) and Ko-jumdgieva et al. (1989).

During the Lower Miocene Northwest Bulgariawas an upland area and no sediments are knownfrom this time period. The beginning of the Mid-dle Miocene is characterised by a large marinetransgression £ooding most of Northwest Bulga-ria which becomes part of the Forecarpathian ba-sin. The Miocene longitudinal depression wasformed (Fig. 1, ‘II’), surrounded by the shallowplatform area named the ‘Marginal stable area’(Fig. 1, ‘I’). Due to the existing sea way betweenthe Forecarpathian and Pannonian basins at thistime (Ro«gl, 1998) facies and fossil content of thesediments of both basins are quite similar. Hence,

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^37 23

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^3724

the stratigraphic scheme of the Pannonian basincan also be used in the study area (Kojumdgievaand Popov, 1986; cf. Table 1). At the end of theBadenian the uplift of the Carpathian mountainscaused a signi¢cant regression and before the endof the Badenian the sea retreated from the terri-tory of NW Bulgaria (Table 1, Fig. 2) Possiblerelated to changing sea levels the Badenian isalso characterised by the formation of evaporites.

In the Lower Sarmatian (i.e. Volhynian) a newtransgression occurred. The longitudinal depres-sion and the marginal stable area in the southwere covered by sea water. The extension of theLower Sarmatian basin was approximately thesame as in the Lower Badenian. From the Sarma-tian onwards until the end of the Miocene therewas no connection with the Pannonian basin, buta sea way to the Eastern Parathetys developed(Ro«gl, 1998). Correspondingly, sedimentationand faunal communities of the Forecarpathianand Euxinian basin are quite similar so that thestratigraphic scheme for the Eastern Paratethyscan also be applied to the sediments in NorthwestBulgaria (Kojumdgieva and Popov, 1986;Table 1). During the Sarmatian the subsidenceof the longitudinal depression ceased, and thisarea was almost ¢lled with sediments. This ex-plains why in the cores Makresh C-37 and Dele-ina C-12 sedimentation stops during the Middleand Upper Sarmatian (Bessarabian and Cherso-nian), respectively (Fig. 2). But a new depressionappeared, situated transversally to the former one,namely the ‘Lom depression’ (Fig. 1, ‘III’). Dur-ing the early Chersonian the extension of the ba-sin is quite limited and ¢nally, the sea left theterritory of NW Bulgaria (Table 1, Fig. 2).

Fig. 3. Simpli¢ed pollen diagram of core C-12 Deleina.Shaded area: 20 times exaggerated (after Ivanov, 1995).

Fig. 2. Lithologic columns of the studied pro¢les. The loca-tions of the palyno£oras studied are indicated by bareswhich show the ratio of palaeotropic over arctotertiary ele-ments for the corresponding sample. Legend: (1) clay;(2) sandy clay; (3) siltstone; (4) sand; (5) conglomerate;(6) limestone; (7) gypsum. A1 ^ arctotertiary elements,warm^temperate; A2 ^ arctotertiary elements, cool-temper-ate; A ^ arctotertiary elements which cannot be assigned toA1 or A2; P1 ^ palaeotropic elements, tropical; P2 ^ palaeo-tropic elements, subtropical; P ^ alaeotropical elementswhich cannot be assigned to P1 or P2.

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^37 25

After a time lag, which is not well de¢ned (itprobably encompasses the middle and upper partsof the Chersonian; Fig. 2, Table 1), the nexttransgression occurred in the Lower Maeotianand the basin occupied primarily the Lom depres-sion, its neighbouring areas and a small part ofthe Vidin area. During the Upper Maeotian, thearea covered by the basin shrank slightly in theEast whereas some parts of the former longitudi-nal depression were covered in the west. The Lomdepression was active until the Lower Pontianwhen the uplift of the Carpathians stopped itsactivity and it was ¢lled with sediments. TheUpper Pontian and Lower Pliocene are only rep-resented by continental sediments with a limiteddistribution.

4. Vegetation dynamics

The following description of the Miocene vege-tation dynamics in the Southern Forecarpathianbasin is a brief summary of the analysis of Ivanov(1995). In addition we provide information aboutchanges in the ratio of palaeotropical over arcto-tertiary elements (P/A-ratio) which presumably re-£ects climatic changes in the study area (see Sec-tion 5).

4.1. The Badenian

The Badenian of the Forecarpathian basin issubdivided into three substages, i.e. Lower, Mid-dle, and Upper Badenian (see Kojumdgieva andPopov, 1989; Kojumdgieva et al., 1989; Table 1,Fig. 2). The information about the compositionand characteristic features of the vegetation dur-ing the Badenian is obtained from pollen dia-grams of the cores Deleina C-12 and MakreshC-37 (Figs. 2^4) which cover only the Middleand Upper Badenian. Characteristic for the vege-tation of that time is the regular occurrence andabundance of thermophilous species like Engel-hardia, Reevesia, Theaceae. Pandanus, Symplocos,Sapotaceae, Araliaceae, Arecaceae, Schizaeaceae,Gleicheniaceae. Together with the relatively highP/A-ratio (Fig. 2) this suggests the existence ofwarm, subtropical climate during the Badenian

and a tendency towards slightly cooler conditionsat the end of the period (cf. P/A-ratio in Fig. 2).

Correspondingly, the arctotertiary elements ofthe mixed mesophytic forests (such as Fagus, Al-nus, Carpinus, Betula, Tilia, Tsuga) are less abun-dant than in the overlying sediments and somearctotertiary taxa are very rare (less than 1%)and hence not represented in the diagrams ofFigs. 3 and 4. It should also be mentioned thatconifer pollen can be particularly abundant, pre-sumably because of the capacity of saccate pollenfor long distance transport: during the Badenianthe basin developed its largest extension so thatthe cores have the maximum distance from thecoast line. Correspondingly, the relative abun-dance of the other pollen types and of the P/A-ratio does not change. Thus, the high percentageof conifer pollen does not provide evidence for adominant coniferous vegetation or cooler climaticconditions.

The dominant elements of the mesophytic for-ests are Engelhardia, Quercus, Castanopsis andCastanea. In addition, many subtropical speciesgrew in these forests, e.g. Reevesia, Chloranthus,Itea, Symplocos, Alangium, Arecaceae, cf. Mastix-iaceae, Schizaeaeceae. They indicate humid sub-tropical conditions and some of them presumablyrepresent Palaeogene relics.

The swamp forests were also well developedduring the Badenian. Its components, such asthe Taxodiaceous pollen, Nyssa, Myrica, Planera,show comparatively high values in the pollenspectra. Probably the relief and palaeogeographicsituation of that time favoured the wide distribu-tion of swamp forests and of ecologically relatedriparian forests with Platanus, Planera, Liquidam-bar, Ulmus, Carya, Pterocarya and Salix. Themost favourable conditions for these forest typesexisted in the early Upper Badenian as is evidentfrom the particularly high percentages of repre-sentatives of these vegetation types between473.0 and 459.0 m of core C-12 (Fig. 3; this phaseis not represented in core C-37, see Fig. 2). It isalso interesting to note that high percentages ofTaxodiaceae correlate with high frequencies ofNyssa, Myrica, Platanus, Planera, Liquidambar,Carya, Pterocarya, and Salix.

The development of favourable conditions for

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^3726

swamp forests in the early Upper Badenian ofcore C-12 was probably linked with the beginningof a regression and the successive isolation of theForecarpathian Basin (see above). Trashliev(1984) discussed the various stages in the forma-tion of Badenian evaporites. He argued that as aresult of an interrupted water exchange with thesea, the Forecarpathian Basin was transformedinto a closed basin at the end of the Middle Ba-denian so that evaporites could precipitate in ba-sin lowlands and lagoons. At the end of this ‘la-goon period’, a slight increase of the sea level incombination with an increase of precipitation andair humidity led to the restoration of water in£owinto the basin: the gypsum deposition ceased andfresh water conditions reappeared. Namely theseenvironmental conditions caused the optimum de-velopment of swampy vegetation in sediments im-mediately above the gypsum layer. After this re-storation of the connections with the sea the basinoccupied throughout the rest of the Badenian al-most the same territories as in the Middle Bade-nian. Finally, near the end of the Badenian thesea retreated from the territory of NW Bulgaria ina major regression.

4.2. The Sarmatian

The Sarmatian of the Forecarpathian basin issubdivided into three substages, i.e. Volhynian,Bessarabian, and Chersonian, also termed Lower,Middle and Upper Sarmatian (see Kojumdgievaand Popov, 1989; Kojumdgieva et al., 1989). It isa period when a rich and diverse vegetation de-veloped and shows a tendency towards increasingproportions of arctotertiary elements such as Be-tula, Alnus, Carpinus, Corylus, Fagus, Eucommia,and Tilia while typical subtropical elements likeEngelhardia, Alangium Reevesia, Itea, Pandanus,Castanopsis, Sapotaceae, Symplocaceae, Thea-ceae, Arecaceae tend to decrease in abundance(Figs. 3^5). This paleo£oristic change occursslowly and gradually without major £uctuations.Correspondingly the P/A-ratio (Fig. 2) shows mi-

Fig. 4. Simpli¢ed pollen diagram of core C-37 Makresh.Shaded area: 20 times exaggerated (after Ivanov, 1995).

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^37 27

nor £uctuations during the Lower Sarmatian andthen a distinct trend towards lower values duringthe rest of the Sarmatian. A similar vegetationchange is observed in the Sarmatian of otherareas of the Eastern and Central Paratethys (e.g.Ukraine: Syabryay and Stekina, 1983; Slovakia:Planderova¤, 1990; Hungary: Nagy, 1992).

The Volhynian vegetation retains many featuresof the Upper Badenian and most favourable con-ditions existed for the development of mixed mes-ophytic forests which are characterised by a pre-dominance of warm^temperate and subtropicalelements. Correspondingly, many palaeotropicalelements are still present in the pollen spectra.During the Volhynian, the structure of the meso-phytic forests changes and some arctotertiary spe-cies become more abundant. For instance, Tsugais absent in the Badenian or very rare, it is stillrare at the base of the Volhynian, but reachespercentages up to 10% in the middle and upperpart of the Volhynian (Figs. 3 and 4).

The distribution of swamp forests seems to bemore restricted in the Sarmatian as compared tothe Badenian. The percentage of Taxodiaceae ismostly lower and the genera Myrica, Nyssa areless abundant. There is only a short period inthe Volhynian which is characterised by a highabundance of this vegetation type (Fig. 3: 264^258 m). In the eastern part of the basin, near thevillages of Stavertsi and Staroseltsi (Pleven dis-trict), Atanassov et al. (1971) observed the sameshort expansion of swamp vegetation during themiddle part of the Volhynian.

The basal parts of the Bessarabian are charac-terised by a vegetation similar to the Volhynianone, but at the end of this substage the abundanceof palaeotropical elements tends to decrease as isevident from the pollen spectra and the P/A-ratio(Fig. 2). The swamp forests existed without majorchanges but towards the end of the Bessarabianthe Chenopodiaceae become more abundant(Figs. 3 and 4). This may be due to a somewhatdryer climate during that time as is also indicatedby macrofossil data (Palamarev, 1991; Palamarev

Fig. 5. Simpli¢ed pollen diagram of core C-12 Drenovets.Shaded area: 20 times exaggerated. (After Ivanov, 1995).

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^3728

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^37 29

and Ivanov, 2001). This trend continued duringthe Chersonian when the Chenopodiaceae reachup to 16%. Presumably open landscapes coveredby more xerophytic herbaceous communities ex-isted during that time. As regards the mesophyticforests, Quercus, Ulmus, Carya, and Pterocaryaare the dominant elements while the pollen ofSymplocos, Reevesia, Itea, Sapotaceae and otherthermophilous species are absent. Abies and Piceapollen also occur with higher percentages.

As mentioned above, the characteristic featureof the Bessarabian is the expansion of Chenopo-diaceae which may re£ect an increasingly drierclimate, a process which started already in theBessarabian and is con¢rmed by sedimentologicaldata (Koleva-Rekalova, 1994, Ivanov and Kole-va-Rekalova, 1999): in Bessarabian to Chersoniansediments of North-East Bulgaria aragonite sedi-ments occur which are assumed to have beenformed under a seasonally dry climate. The exis-tence of drier conditions at the end of the Sarma-tian (s. 1.) is also established in other areas ofthe Balkan peninsula (Pantic and Mihailovich,1980).

4.3. The Maeotian and Lower Pontian

During this time period, which is only repre-sented in core Drenovets C-1 (Figs. 2 and 5),Quercus, Castanea, Ulmus, and Carya are thedominant elements of the mesophytic forests.Moreover, Fagus, Betula and Carpinus are con-stantly present. Coniferous forests also played amore signi¢cant role as documented by increasedpercentages of Pinus, Abies, Picea and Cedrus :thereby the increase is particularly evident forthe Pinus haploxylon type. Sporadically pollengrains of Ilex, Eucommia, Hedera, Parthenocissus,Magnolia, Reevesia, Itea, Sapotaceae, Araliaceaeand Arecaceae occur with percentages below 1%.The other thermophilous elements like Engelhar-dia are also present in low quantities. Among thepteridophytes Schizaeaeceae and Cleicheniaceae

are missing. In other areas of Bulgaria Palamarev(1991) and Palamarev et al. (1999) observed asimilar tendency towards a reduction of thermo-philous elements of the families Lauraceae, Ju-glandaceae, Araliaceae and Staphyleaceae duringthe Maeotian. In addition, the P/A-ratio is lowestin the Maeotian (Fig. 2) showing only some mi-nor £uctuations but a distinct increase occurs inthe Lower Pontian.

The swamp forests show a restricted distribu-tion in the Maeotian. Taxodiaceae are less abun-dant, the genera Nyssa and Myrica occur sporadi-cally and in very low numbers; whereas thefrequency of Alnus pollen increases. The same istrue for the riparian forests the main componentsof which are also relatively rare except for Salix.However, riparian and swamp forest elements be-come more abundant at the end of the Maeotianand in the Lower Pontian (cf. Alnus, Salix, Nyssain Fig. 5).

The herbaceous communities have their maxi-mum distribution in the Maeotian. Besides Che-nopodiaceae, Artemisia, Caryophyllaceae, Polyga-laceae, Lamiaceae, Asteroideae, Cichorioideae arealso present. Then, during the latest Maeotianand Lower Pontian there is again a decline ofmost herbaceous elements (this is not true, how-ever, for the Poaceae which may grow in verydi¡erent environments). This fact as well as theincrease of riparian and swamp forests may indi-cate more humid conditions at that time.

5. Reconstruction of Palaeoclimate dynamics

The paleaoclimate reconstructions resultingfrom the application of the coexistence approachto the palyno£oras of cores C-12, C-37 and C-1are shown in Figs. 6^8, respectively. The present-day climate of Northwest Bulgaria is character-ised by MAT 11.2^11.5‡C, TCM 32.1 to30.9‡C, TWM 22.6^23.6‡C and MAP 536^586mm (Velev, 1997).

Fig. 6. Coexistence intervals and P/A-ratio derived from samples of core Deleina C-12. MAT ^ mean annual temperature; TCM^ mean temperature of the coldest month; TWM ^ mean temperature of the warmest month; MAP ^ mean annual precipitation;P/A-ratio ^ ratio of palaeotropic over arctotertiary elements.

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^3730

Fig. 7. Coexistence intervals and P/A-ratio derived from samples of core Makresh C-37 (abbreviations: see Fig. 6).

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^37 31

5.1. The Badenian

The data obtained from cores Deleina C-12(Fig. 6) and Makresh C-37 (Fig. 7) display rela-tively stable climatic conditions for the Badenian.The MAT coexistence intervals range mainly be-tween 15.6^18.4‡C but intervals between 17.2^18.4‡C also occur; TCM is between 5.0 and8.1‡C or even 9.4‡C, and the TWM coexistenceinterval is constantly 24.7^27.8‡C. The lowerboundary of MAP coexistence intervals lies at823, 1076 or 1187 mm, and the upper boundaryat 1308 or 1322 mm with the most common in-terval being 1187^1308 mm. Thus, all data indi-cate a very warm and humid subtropical climate.At the end of the Badenian a slight decrease inMAT and TCM is observed in core C-12, as ex-pressed by decreasing upper boundaries of thecoexistence intervals. This corresponds with de-creasing P/A-ratios and changes in plant com-munities discussed above in the chapter describingthe Badenian vegetation. Presumably all thesechanges re£ect the transformation of the paleo-geographic situation at that time, when sea regres-sion took place thus inducing some drying andslight cooling of winter temperatures: the climatebecomes more continental with an increased sea-sonality. In core C-37 this period of cooler con-ditions with reduced P/A-ratios is not represented,presumably because sedimentation stopped earlierthan in core C-12 (cf. Section 2).

5.2. The Sarmatian

In the Sarmatian the climatic conditions stillremain relatively stable during the Volhynianand the greater part of the Bessarabian (Fig. 6and 8). MAT is between 15.6 and 17.2‡C,although some samples give higher temperaturesand a second coexistence interval 17.2^18.4‡C,may appear (Fig. 6); for an explanation of theoccurrence of a second coexistence interval seeSection 2. TCM is mainly between 5 and 7‡C,summer temperatures lie within 24.6^27.8‡C andboth TCM and TWM show small oscillations ofthe upper limit. As regards the £uctuations ofMAP, the narrowest coexistence intervals occurduring the upper part of the Volhynian with

1187^1322 and 1076^1322 mm; this possibly cor-responds to the highest precipitation rate in theSarmatian as is also indicated by the vegetationdata (see chapter describing the Sarmatian vege-tation). The P/A-ratio remains largely constantshowing only minor £uctuations which may re-£ect changes in temperature and humidity.

In terms of palaeoclimate, the Bessarabian ismost similar to the Volhynian. However, at theend of this substage a slight decrease in MAT isindicated for core C-1. Meanwhile, the P/A-ratiodecreases from ca. 35% to values between 25%and 20% in cores C-12 and C-1. Thus the endof the Bessarabian represents the starting pointfor the climatic changes occurring in the Cherso-nian. This latter substage is characterised by lowervalues in almost all climate parameters (Fig. 6).MAT coexistence intervals are 13.3 (14.4) to17.2‡C, corresponding to a decrease of the lowerboundary of about 2‡C as compared to the Bes-sarabian and Volhynian. A similar cooling is ob-served for the TCM with the lower boundary ofthe coexistence intervals changing from 5‡C in theBessarabian/Volhynian to 1.7‡C in the Cherso-nian. The lower boundary of the coexistence in-terval for the summer temperature (TWM) alsodecreases by about 2‡C. But the most signi¢cantchange occurs in the MAP. The coexistence inter-vals for MAP fall from ca. 1076^1308 mm to 652(740)^759 mm although a second, more humidcoexistence interval may appear (Fig. 6; for aninterpretation of the second coexistence intervalsee description of the Maeotian climate below).These climatic estimates for the Chersonian indi-cate that the climate was slightly cooler (particu-larly in winter) and signi¢cantly drier than in theBessarabian which is also evident from the vege-tation change described in a previous chapter.Similar vegetation changes, caused by a dryerand cooler climate, occurred throughout the Para-tethyan realm. It should be noted that in the areaof Slovakia Bruch (1998) found a similar MAPcoexistence interval of ca. 750^760 mm for thistime interval. There is some evidence that thesechanges were not equal in the northern and south-ern parts of the Paratethys and that the dryingwas more intense in the South (Ivanov and Kole-va-Rekalova, 1999). Probably, the distribution of

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^3732

Fig. 8. Coexistence intervals and P/A-ratio derived from samples of core Drenovets C-1 (abbreviations: see Fig. 6).

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^37 33

Fig. 9. Summarised data of coexistence intervals and P/A-ratios for the Badenian to Pontian (Middle^Upper Miocene). In addi-tion to the coexistence intervals the curve linking their middle values is shown. When there is more than one coexistence intervalper sample, the middle value of all intervals has been calculated. The curves of the middle values do not imply that they repre-sent the ‘real’ values, but they express quite well the overall trends of climate change.

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^3734

land and water basins played a major role in de-termining local climate and precipitation. Shrink-ing water bodies caused greater seasonality andthe development of a dry period, which was prob-ably the limiting factor for the distribution ofplants and plant communities. The upper partof the Chersonian is not documented becausethe sea abandoned the territory of Northwest Bul-garia.

5.3. The Maeotian and Lower Pontian

Palaeoclimatic parameters calculated for thistime period display increased climate £uctuationsand variations of the P/A-ratio (Fig. 8). For someof the analysed samples we obtained two MAPcoexistence intervals as was the case for someChersonian samples (for an explanation of thisphenomenon see Section 2 and Mosbrugger andUtescher, 1997). The ¢rst coexistence interval ap-pears at ca. 740^760 mm and the second one atca. 900^1300 mm. Possibly these two coexistenceintervals represent two di¡erent plant commun-ities which grow under di¡erent climatic condi-tions depending on the variegated relief. The ¢rstcoexistence interval may represent the climaticconditions of the lowlands and open areas sur-rounding the basin where herbaceous commun-ities developed growing under a dryer climate.The second coexistence interval is determined bythe components of mixed mesophytic forests.Here it is important to keep in mind that moun-tains existed near the basin, approximately lessthan 20 km to the South. These mountains arethe precursors of the present-day Balkan moun-tains which in the study area are up to 2000 mhigh. The presence of high mountains near a largewater basin usually implies accumulation of highair humidity and more rainfall in the mountainarea. Thus mixed mesophytic forests may havefound suitable climatic conditions on, and nearthe Palaeo^Balkan mountains. Moreover, the dif-ference between the two coexistence intervals isonly ca. 150^200 mm and it is reasonable to as-sume that the precipitation was at least 200 mmhigher in the mountain area.

Looking in more detail at the quantitative cli-mate data we see that the Lower Maeotian begins

with a somewhat warmer and more humid climatethan existed during the Chersonian with MAT =15.6^17.2‡C, TCM = 5.0^6.6‡C, TWM = 24.7^27.3‡C and MAP up to 1187^1322 mm. This earlyphase is followed by a cooler period: the lowerboundary of MAT decreases by about 2^3‡C, andthe lower boundary of TCM reaches 0‡C. TWMalso shows a decreasing trend as expressed by thecurve of the middle values of the coexistence in-tervals (Fig. 9). After that, in the middle of theUpper Maeotian, we observe a new oscillationtowards warmer (MAT, TCM) and more humidconditions (Figs. 8 and 9). This period is followedby a dryer and cooler phase at the end of theMaeotian; here we observe the lowest MAT coex-istence intervals with the lower boundary beingbelow 12‡C.

After the Upper Maeotian cooling, the Pontianbegins with a warming trend. MAT reaches valuesof 15.6^17.2‡C, TCM is 5^7‡C. Almost simulta-neously MAP increases up to 1187^1308 mm.This event corresponds with the vegetation datareconstructed in a previous chapter which showan expansion of swamp and riparian forests dur-ing the Lower Pontian.

Summarising the data for the Maeotian^LowerPontian we can conclude that the climate was stillwarm^temperate with several oscillations in tem-perature and precipitation. The changes in theP/A-ratio follow approximately the changes inMAP. Presumably, during this period vegetationdynamics, including the abundance of palaeotrop-ical elements, was largely controlled by humidityand rainfall. This is in contrast to previous peri-ods when temperature also played a major role incontrolling the distribution of these elements.

6. Conclusions

The well-preserved and diverse palyno£orasfrom Northwest Bulgaria (Southern part of theCentral Paratethys) allowed us to reconstruct veg-etation and climate evolution and to link it withpalaeogeographic changes. With the help of thecoexistence approach we obtained the ¢rst quan-titative climatic reconstruction for the entirestratigraphic sequence from the Middle to Upper

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^37 35

Miocene in the Southern part of the Paratethyanarea, summarised in Fig. 9.

The middle Miocene (Badenian^Volhynian ^greater part of Bessarabian) was a period of sub-tropical/warm^temperate and humid climate. Thedata reveal only small £uctuations in the palaeo-climatic parameters (Fig. 9). The palaeogeo-graphic reorganisation and regression led to de-creasing MAT and MAP at the end of theBadenian. Over the whole period the temperaturestend to decrease but probably not more than 1^2‡C. Simultaneously, the ratio between paleotrop-ical and arctotertiary elements also decreases. Theclimate and vegetation data show a step-by-steptransformation without extreme deviations. Dur-ing the Upper Miocene (Chersonian to Pontian)diverse climatic conditions occurred and in£u-enced the vegetation development (Fig. 9). Thisvariation was probably not only due to changesin the paleogeographic situation, but also to glob-al or regional climatic changes. The beginning ofthe Upper Miocene is marked by a slight coolingand a signi¢cant drying of the climate. The samecooling trend is observed in other areas of Eu-rope, e.g. in Northwest Germany (Gebka et al.,1999; Utescher et al., 2000). Later, £uctuationsoccur in all climate parameters and display cyclesof humid/dryer and warmer/cooler conditions.These oscillations of MAT, TCM and TWMwere probably within the limits of 3^4‡C; togeth-er with the changes in precipitation they were thefactors in£uencing vegetation dynamics.

Our palaeoclimate analysis of the Middle toUpper Miocene Forecarpathian basin provides arelatively high resolution with respect to stratig-raphy but also with respect to the precision of theclimate reconstruction. As is evident from Fig. 9,we cover a time span of about 8 Ma with 69 timehorizons and provide quantitative data for fourdi¡erent palaeoclimate parameters. Thus, the re-sults presented above can be used as a ¢rst modelof the climate evolution during the Middle^UpperMiocene in Southeast Europe.

Acknowledgements

The financial support of the Deutsche Akade-

mische Austauschdienst, of the Deutsche For-schungsgemeinschaft and of the CollaborativeResearch Center 275 (Tu«bingen) are gratefullyacknowledged. This publication is a contributionto the international NECLIME project (www.ne-clime.de). We would like to thank Mike Boulter,London, and one anonymous reviewer for criti-cally reading the manuscript and for their valuablesuggestions. James Nebelsick, Tu«bingen, helped tocorrect the English.

References

Atanassov, G., Popov, N., Trashliev, S., Kojumdgieva, E.,Goranov, A., Dakov, D., 1971. Lithostratigraphical and fa-cial characterization of Sarmatian near Pleven (in Bulgarian,English abstr.). Ann. Univer. So¢a Geol. 63 (1), 181^216.

Bruch, A., 1998. Palynologische Untersuchungen im Oligoza«nSloweniens-Pala«o-Umvelt und Pala«oklima im Ostalpenraum.Tu«b. Mikropala«ontol. Mitt. 18, 193S.

Gebka, M., Mosbrugger, V., Schilling, H.-D., Utescher, T.,1999. Regional-scale palaeoclimate modeling on softproxy-data basis ^ an example from the Upper Miocene ofthe Lower Rhine Embayment. Palaeogeogr. Paleoclimatol.Paleaeoecol. 152, 225^258.

Hadgiev, P., Palamarev, E., 1962. Sarmatian £ora in NorthBulgaria. I. Michailovgrad region (in Bulgarian, Germanabstr.). Mitt. Bot. Inst. 10, 5^13.

Ivanov, D. 1995. Palynological investigation of Miocene sedi-ments from North-West Bulgaria (in Bulgarian, Englishabstr.). Autoref. Ph. D. Thesis, Inst. Bot. BAS, So¢a, 45 p.

Ivanov, D., 1997. Miocene palynomorphs from the Southernpart of the Forecarpathian Basin (Northwest Bulgaria).Flora Tert. Mediterr. VI (4), 81.

Ivanov, D., Koleva-Rekalova, E., 1999. Palynological and sed-imentological data about Late Sarmatian paleoclimaticchanges in the Forecarpathian and Euxinian basins (NorthBulgaria). Acta Palaeobot., Suppl. 2. Proceedings 5th EPPC,pp. 307^313.

Kojumdgieva, E., Popov, N., 1986. Stratigraphy and correla-tion of the Neogene in Mediterranean area: results and dis-cussion (in Bulgarian). Rev. Bulg. Geol. Soc. 47 (3), 124^126.

Kojumdgieva, E., Popov, N., 1988. Lithostratigraphy of theNeogene sediments in Northwestern Bulgaria (in Bulgarian,English abstr.). Palaeontol. Stratigr. Lithol. 25, 3^26.

Kojumdgieva, E., Popov, N., 1989. Pale¤oge¤ographie et e¤volu-tion ge¤odynamique de la Bulgarie Septentrionale au Ne¤o-ge'ne. Geol. Balcan. 19 (1), 73^92.

Kojumdgieva, E., Popov, N., Stancheva, M., Darakchieva, S.,1989. Correlation of the biostratigraphic subdivision of theNeogene in Bulgaria after molluscs, foraminifers and ostra-cods (in Russian, English abstr.). Geol. Balcan. 19 (3), 9^22.

Koleva-Rekalova, E., 1994. Sarmatian aragonite sediments in

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^3736

North-eastern Bulgaria ^ origin and diagenesis.. Geol. Bal-can. 25 (5), 47^64.

Mai, D.H., 1995. Tertia«re Vegetationsgeschichte Europas. G.Fischer, Jena, 563 S.

Mosbrugger, V., 1995. New Methods and Approaches in Ter-tiaty Palaeoenvironmental Research. Abh. Staatl. Mus. Min-eral. Geol. Dresden 41, 41^52.

Mosbrugger, V., Utescher, T., 1997. The coexistence approach^ a method for quantitative reconstruction of Tertiaty ter-restrial palaeoclimate using plant fossils. Palaeogeogr.Palaeoclimatol. Palaeoecol. 134, 61^86.

Nagy, E., 1992. A comprehensive study of Neogene sporo-morphs in Hungary. Geol. Hung. Ser. Palaeontol. 53, 376^379.

Palamarev, E., 1988. Sche¥era chandlerae sp.n., a new sub-tropical element in the Bulgarian Neogene Flora. Tert.Res. 9 (1^4), 97^106.

Palamarev, E., 1989. Palaeobotanical evidences of the Tertiaryhistory and origin of the Mediterranean sclerophyll dendro-£ora. Plant Syst. Evol. 162, 93^107.

Palamarev, E., 1990. Grundzuge der pala«o£oristischen Pala«o-sukzessionen im Spa«tmioza«n (Sarmatien-Pontien) Bulga-riens. In: Paleo£oristic and paleoclimatic changes in theCretaceous and Tertiary, Prague, pp. 257^263.

Palamarev, E. 1991. Composition, structure and main stages inthe evolution of Miocene paleo£ora in Bulgaria (in Bulgar-ian). Autoref, DSc thesis, BAS, So¢a, pp. 1^60.

Palamarev, E., 1993. Uº ber die Tetria«re Geschichte der GattungJuglans L. in Bulgarien. Acta Palaeobot. 33 (1), 299^307.

Palamarev, E., Ivanov, D., 2001. Charakterzu«ge der Vegeta-tion des Sarmatien (Mittel- bis Obermioza«n) im su«dlichenTeil des Dazischen Beckens (Su«dost Europa). Palaeonto-graphica 259 (1-6), 209^220.

Palamarev, E., Ivanov, D., Bozukov, V., 1999. Pala«o£oren-komplexe im Zentralbalkanischen Raum und ihreEntwicklungsgeschichte von der Wende Oligza«n/Mioza«n bisins Villafranchien. Flora Tert. Mediterr. VI (5), 95.

Palamarev, E., Petkova, A., 1987. La Macro£ore du Sarmatian(in Bulgarian, French abstr.).. Foss. Bulg. VIII ((1)), 275.

Palamarev, E., Petkova, A., Uzunova, K., 1975. FossileFarnp£anzen aus dem Mioza«n Bulgariens. Fitologija 2,25^33.

Palamarev, E., Petkova, A., Uzunova, K., 1978. A contribu-tion to the history of genera Taiwania Hay. and Cunning-hamia R. Br. in the Holarctic (in Bulgarian, German abstr.).Fitologija 9, 3^16.

Palamarev, E., Uzunova, K., 1970. Morphologisch-anato-

mischer Nachweis der Gattung Skimmia in der Tertia«r£oraBulgariens. C. R. Acad. Bulg. Sci. 23 (7), 835^838.

Palamarev, E., Uzunova, K., 1992. Beitrag zur Entwicklungs-geschichte der Cycadeen in der Tertia«r£ora Europas. Cour.Forsch. Inst. Senckenberg 147, 287^293.

Pantic, N., Mihailovich, D., 1980. Neogene £orae of ‘Balkanland’ Balkan and their signi¢cance for paleoclimatology,paleobiography and biostratigraphy. Ann. Geol. Penin. Bal-kan. XLIII^XLIV, 239^261.

Petkova, A., 1967. Pala«obotanische Untersuchung des Sarmatsan der unteren Stro«mung des Iskar-Flusses (in Bulgarian,German abstr.). Mitt. Bot. Inst. 17, 135^168.

Petkova, A., 1977. The Sarmatian £ora from Northwest Bul-garia: taxonomic, phytogeographic and palaeoecologicalanalysis (in Bulgarian). PhD thesis, Inst. Bot., BAS, 286 pp.

Petkova, A., Kitanov, B., 1965. Sarmatische Flora aus demBezirk Vidin, Nordwest-Bulgarien (in Bulgarian, Germanabstr.). Ann. Univ. So¢a Biol. 59 (2), 1^25.

Planderova¤, E., 1990. Miocene micro£ora of Slovak CentralParatethys and its biostratigraphical signi¢cance. DionizStur Institute of Geology, Bratislawa, 144 pp.

Ro«gl, F., 1998. Palaeogeographic Considerations for Mediter-ranean and Paratethys Seaways (Oligocene to Miocene).Ann. Naturhist. Mus. Wien 99A, 279^310.

Stefanov, B., Kitanov, B., Petkova, A., 1964. Une trouvaille demate¤riaux fossiles the Metasequioia gliptostroboides Hu etCheng dans le Mioce¤ne de la Bulgarie du Nord (in Bulgar-ian, French abstr.). Ann. Univ. So¢a Biol. 58 (2), 1^8.

Syabryay, S.V., Stekina, N.A. 1983. History of the develop-ment of the vegetation of Ukraine during the Miocene (inRussian). Naukovaya Dumka, Kiev, 166 pp.

Trashliev, S., 1984. Stages in the formation of Badenian evap-orites North-East of Vidin (in Bulgarian, English abstr.).Palaeontol. Stratigr. Lithol. 20, 76^84.

Utescher, T., Mosbrugger, V., Ashraf, A.R., 2000. TerrestrialClimate Evolution in Northwest Germany Over the Last 25Million Years. Palaios V (15), 430^449.

Uzunova, K., 1995. Dispersed cuticles of Sarmatian sedimentsin Northwestern Bulgaria. Lauraceae. Phytol. Balc. 1 (1),13^17.

Uzunova, K., 1996. Dispersed cuticles from the Sarmatiansediments in Northwestern Bulgaria. Phytol. Balc. 2 (1),29^36.

Velev, S. 1997. Contemporary air temperature and precipita-tion £uctuations in Bulgaria. In. Jordanova, M., Donchev,D. (Eds.), Geography of Bulgaria (In Bulgarian). Acad.Publ. House, So¢a, pp. 145^150.

PALAEO 2724 11-4-02

D. Ivanov et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 19^37 37