Vegetation history in the Eastern Romanian ... - Springer Link

1 downloads 0 Views 1MB Size Report
Aug 5, 2003 - Vegetation history in the Eastern Romanian Carpathians: pollen analysis of two sequences from the Mohos¸ crater. Received: 16 May 2001 ...
Veget Hist Archaeobot (2003) 12:113–125 DOI 10.1007/s00334-003-0015-6

ORIGINAL

Ioan Tantau · Maurice Reille · Jacques-Louis de Beaulieu · Sorina Farcas · Tomasz Goslar · Martine Paterne

Vegetation history in the Eastern Romanian Carpathians: pollen analysis of two sequences from the Mohos¸ crater Received: 16 May 2001 / Accepted: 10 June 2003 / Published online: 5 August 2003  Springer-Verlag 2003

Abstract Two sequences of about 10.5 m originating from a peat bog in Romania were analysed for pollen (202 and 127 pollen spectra). The vegetation history, supported by 24 14C dates is described since the Late Glacial. At the onset of the Holocene Ulmus first appears, together with Betula. Among the main components of the Quercetum mixtum (Quercus, Fraxinus, Tilia, Corylus) that became established almost simultaneously by around 9000 B.P., Quercus frequencies rarely exceed 10%. The local establishment of Carpinus is about 6000 b.p. Its maximum occurred between 4500 and 3000 b.p. Fagus pollen is regularly recorded since 8000 b.p. Its absolute dominance tooks place at about 3000 b.p. Picea pollen is present since the Late Glacial. The first indications of human activities appear at around 6500 b.p.

Introduction The present study is a contribution to a programme of systematic reinvestigation of key sites in Romania aiming to establish an absolute chronology of the vegetation history since the Late Glacial. It follows a first article (Farcas et al. 1999) in which the first 14C dates relating to the vegetation history of the Carpathian mountains were presented. The Mohos¸ peat bog (25550 E; 46050 N; 1,050 m altitude) is situated in the elbow formed by the Carpathian mountains (Fig. 1), in the Ciomadul massif, near TusnadBai. This vast and beautiful Sphagnum peat bog occupies

Keywords Romania · Vegetation history · Late glacial · Holocene

I. Tantau · M. Reille · J.-L. de Beaulieu ()) Facult des Sciences, LBHP IMEP (UMR CNRS 6116), St. Jrme, 13397 Marseille, France e-mail: [email protected] Tel.: +33-442-908477 Fax: +33-491-288668 I. Tantau Department of Geology, University Babes-Bolyai, M. Koga˘lniceanu Street 1, 3400 Cluj-Napoca, Romania S. Farcas Institute of Biological Researches, Republicii Street 48, 3400 Cluj-Napoca, Romania T. Goslar Poznan Radiocarbon Laboratory, Ul. Rubiez 45, 61-612 Poznan, Poland M. Paterne quipe Radiocarbone, Laboratoire des Sciences du Climat et de l’environnement, Avenue de la Terrasse, BP 1, 91198 Gif sur Yvette Cedex, France

Fig. 1 Location map of the cited sites: 1 Mohos (1,050 m), 2 Lucs (1,080 m), 3 Iezerul Caliman (1,650 m), 4 Taul Zanogutii (1,840 m), 5 Capatina (1,220 m)

114

Fig. 2 Location map of the Mohos¸ and St. Ana craters (modified after Juvign et al. 1994)

an area of about 80 ha at the bottom of a circular volcanic crater. It is situated at the limit of the beech forest and the spruce forest that covers the summits around the crater. The peat bog itself is partly covered with Pinus sylvestris and Betula pubescens, whereas vegetation at its periphery consists mainly of Alnus glutinosa, Betula pendula and diverse Salix species. At the centre of the peat bog there are various species of Ericaceae (Empetrum nigrum, Vaccinium vitis idaea, Vaccinium oxycoccos, Andromeda polifolia), associated with Sphagnum. The local vegetation type is assigned to the ‘Vaccinio uliginosi – Pinetum silvestri’ plant community, while the specific vegetation stage for this altitude (middle mountainous) is typically characterised by the ‘Symphyto cordati–Fagetum’ plant community (Coldea 1991). A map and a detailed description of the vegetation at the site are presented in an earlier work devoted to the study of peat sites in Romania (Pop 1960). One of the most notable features of this peat bog is the existence of some ten water holes, several meters in diameter and depth, with very steep borders (see map in Pop 1960). The Mohos¸ crater is some 2 km distant from the Sfnta Ana crater (Fig. 2), the latter being occupied by a lake. The perfectly circular perimeter of the Mohos¸ crater suggests that it is an explosion crater. Considering its large diameter, a sediment filling much thicker than the 10-m-thick peat found from the periphery to the centre of the peat bog would be expected. It is therefore likely that the Mohos¸ crater is much older than the Sfnta Ana eruption and that an older Pleistocene sediment filling was sealed by the Late Glacial tephra. However, this hypothesis could not be verified because the handoperated samplers at our disposal could not penetrate the grey clay. The region is characterised by a temperate-mountainous climate with an annual precipitation level above 1,000 mm. Precipitation reaches a maximum during summer and a minimum during winter. The mean annual temperature is 3 C and the mean summer temperature is 15 C.

From the geological point of view, our study site belongs to the long volcanic range Ca˘limani-GurghiuHarghita that extends over 200 km from north to south of the Eastern Carpathian mountains. Ka/Ar datings attribute ages ranging from 9 to 0.85 million years to most of these volcanoes. However, according to Juvign et al. (1994), the volcanic eruptions in the two craters described occurred within a rather short time interval. The eruption of Sfnta Ana would be dated to about 10,700 B.P. It ejected and deposited in the Mohos¸ crater a lapili and ash tephra layer on which the peat sequence accumulated. The oldest deposits of that sequence therefore could not be older than 10,700 B.P. The authors indicate that a sediment sample collected at “l’extrÞme base d’une carotte de tourbe du cratre de Mohos¸ surmontant immdiatement la tephra du Sfnta Ana” ( 4 m de profondeur environ) a donn un ge de 7,610 b.p.” (Juvign et al. 1994). Pop and Diaconeasa (1967) published pollen analyses of two peat sequences from Mohos¸, one 10.1 m deep and the other 6.8 m. These two sequences revealed a similar vegetation history beginning at the onset of the Holocene. The salient features of the forest development are described but no mention is made of Fraxinus, which is one of the major trees at the regional level today. The non arboreal taxa for which a curve is recorded in good detail are Ericaceae and Sphagnum. The other herbaceous taxa were included in the NAP curve.

Material and methods Six corings made by Pop and Diaconeasa (1967) from the centre of the crater to the periphery reached the following depths: 10.5, 10, 15, 9.50, 9.40 and 6.80 m respectively. Juvign et al. (1994) report that, in the western part of the crater, “ plus de 20 m du bord, l’paisseur de la tourbe est pratiquement constante (environ 5 m) sur au moins 100 m de longueur”. Our two borings were made using a hand operated Russian sampler providing cores 8 cm in diameter. The first boring (Mohos¸ 1), made at the centre of the peat bog, reached the bottom at 10.6 m. The second boring (Mohos¸ 2), at about 30 m from the north-western edge of the bog, reached the bottom at 10.2 m. The lowermost part of the two cores sampled consists of limnic grey clay. In the Mohos¸ 1 sequence, a water pocket between 440 and 490 cm interrupts the regularity of the peat deposit. Such phenomena were also reported by Pop (1960). They probably correspond to water channels formed within the peat most likely related to the presence of the above-mentioned water holes. The cores were described in the field, placed in half sections of PVC tube, and wrapped in plastic film. Before sub-sampling, all the cores were carefully cleaned and described again. A lithological description of the cores by G. Digerfeldt is given in Table 1. The cores were regularly sub-sampled at 5 cm intervals for pollen analysis. The sample preparation (1 cm3) followed the standard procedure: acetolysis in the case of peat and gyttja samples and flotation with Thoulet liquid (Goeury and Beaulieu 1979) for clayey samples. Microscope slides were prepared from the residue and examined for pollen. At least 300–350 grains of trees pollen were counted for each sub-sample, except in cases where pollen concentration was low. The nomenclature for vascular plants follows Flora Europaea (Tutin et al. 1964–80). The sequence Mohos¸ 1 (M1) was analysed as 202 pollen spectra, and Mohos¸ 2 (M2) as 127 pollen spectra. The pollen spectra of the Mohos¸ bog were graphically represented using the Gpalwin software created in the Historical Botany

115 Table 1 Stratigraphic description of the sequences (by G. Digerfeldt) MOHOS¸ 1 0–235 cm 235–405 cm 405–455 cm 455–500 cm 500–608 cm 608–743 cm 743–780 cm 780–965 cm 965–984 cm 984–1066 cm 1066–1085 cm MOHOS¸ 2 0–225 cm 225 –335 cm 335–1006 cm 1006–1020 cm >1020 cm

Sphagnum peat, slightly humidified, brown–light brown. Rather successive boundary to the following layer (3–4 cm) Sphagnum peat, moderately humidified, brown–dark brown. Fibres of Eriophorum moderately frequent. Rather distinct boundary (1 cm) to the following layer Sphagnum peat, moderately–highly humidified, dark brown. Fibres of Eriophorum rather frequent. Rather distinct boundary (1–2 cm) to the following layer Hiatus (water pocket) Sphagnum peat, moderately–highly humidified, dark brown. Fibres of Eriophorum rather frequent. Rather distinct boundary (1–2 cm) to the following layer Sphagnum peat, moderately humidified, brown–dark brown. Rather distinct boundary (2–3 cm) to the following layer Sphagnum peat, highly humidified, dark brown. Fibres of Eriophorum frequent. Distinct boundary (0,5–1 cm) to the following layer. 755–775 cm: coring artefact but no apparent lithological modifications Sphagnum peat, moderately humidified, brown–dark brown. Fibres of Eriophorum rather frequent. Successive boundary (4–5 cm) to the following layer Carex-Sphagnum peat, the lowest about 10 cm almost pure Carex peat, moderately humidified, brown. Rather distinct boundary (1 cm) to the following layer Carex peat, highly humidified, with some wood fragment, brown–dark brown. Distinct boundary (>0.5 cm) to the following layer Limnic grey clay Sphagnum peat, slightly humidified, light brown–downwards brown. Successive boundary (5–7 cm) to the following layer Sphagnum peat, slightly humidified, brown–dark brown. Some fibres of Eriophorum. Rather successive boundary to the following layer (2–3 cm) Sphagnum peat, slightly humidified, dark brown. Rather rich in fibres of Eriophorum. Rather successive boundary (2–3 cm) to the following layer Sphagnum-Carex peat, downwards changing to pure Carex peat, moderately humidified, brown–dark brown Limnic grey clay

and Palynology Laboratory of Marseille (Goeury 1997). The frequencies of pollen for each taxa were calculated as percentages of the total sum (AP + NAP). For ecological reasons, Cyperaceae were excluded from the pollen sum. In the pollen diagrams (Fig. 3 and 4), pollen values lower than 0.5% are represented by dots.

14

C dates

Sixteen samples from M1 and five from M2 were dated by the conventional method at the Physics Institute of the Silesian Technical University (Gliwice, Poland). Two unreliable dates were rejected. Two levels of the zone M1.8, with two measurements made in each of them, and one level in zone M2.8 considered synchronous with M1.8, were dated by A.M.S. at the Laboratory of Gif-sur-Yvette (France).

Results Nineteen local pollen zones are distinguished. Sequence M1 starts in the Late Glacial with three pollen zones (zones 1–3) that have no equivalent in M2. The local zone 7, directly following the Fraxinus maximum is absent from sequence M1 (see Table 2). Zone M1.8 is a peculiar characteristic of sequence M1. Only a very weak equivalent is recorded in profile M2 (see vegetation history - LPAZ 8). Zone M2.13 is characterised by spectra rich in Fagus, although this tree has not yet achieved its maximum expansion. There is no equivalent event in M1. This

might be explained by an accidental circumstance, for instance the exceptional preservation of a whole flower or a stamen of Fagus (as sometimes occurs) or a coring artefact. However, as this event is recorded in two consecutive spectra, to preserve the objectivity of our data we have to regard zone M2.13 as real, but occurring only in profile M2. In M1 a hiatus (water pocket), probably of short duration, interrupts the regular progression of Fagus throughout pollen zone 15. Uncalibrated 14C ages are given in Table 3. They also appear by the dated levels in the pollen diagrams (Figs. 3 and 4).

Vegetation history LPAZ 1: The spectra from this zone are characterised by Pinus pollen values of about 50% suggesting an open forest dominated by Pinus (Pinus mugo). Betula pollen values of about 10% attest to the local existence of this tree, under conditions that were similar to those prevailing today at the site. Picea pollen, recorded in all spectra with values sometimes exceeding 1%, reflects the existence of this tree in some local glacial refugia. It was also found in other Late Glacial sequences from the Romanian Carpathians: Taul Zanogutii, Iezerul Caliman (Farcas et al. 1999), Gutai (Bj rkman et al. 2002). Among the mesophilous

116 Table 2 Correspondence between the pollen zones in the two sequences

117 Table 3 Uncalibrated radiocarbon dates Sample name MohoS¸ 1: M1 80 M1 120 M1 180 M1 300 M1 395 M1 445 M1 490 M1 535 M1 585 M1 645 M1 763 M1 763 M1 774 M1 774 M1 700 M1 800 M1 900 M1 950 M1 985 M1 1000 Mohos¸ 2: M2 135 M2 375 M2 545 M2 665 M2 780 M2 975

Lab. no.

Age (b.p.)

Gds 9760 Gds 9758 Gds 9765 Gds 9759 Gds 9781 Gds 9782 Gds 9767 Gds 9761 Gds 9783 Gds 9763 100 309 100 477 100 310 100 478 Gds 9770 Gds 9778 Gds 9779 Gds 9780 Gds 9769 Gds 10615

120.3€2.7 540€160 900€150 1600€150 2290€170 1230€490 3310€170 4220€180 5070€220 6230€240 5660€90 5860€90 4760€90 4880€80 6660€250 7840€230 8930€250 8100€280 9150€350 9750€200

Gds 108 Gds 112 Gds 110 Gds 109 100 311 Gds 111

1090€90 2910€90 4550€100 5780€90 7500€100 8740€160

Remarks

Rejected

Coring Coring Coring Coring

artefact artefact artefact artefact

Rejected

arboreal taxa, only Ulmus and Quercus have a fairly regular presence, suggesting a pollen transport from more or less distant origin. This zone can with certainty be assigned to the Late Glacial interstadial. LPAZ 2: In this zone, which corresponds to the Younger Dryas, the arboreal pollen percentages decrease by half. Poaceae and herbaceous steppe species, particularly Artemisia, experience a major expansion, as occurred everywhere in southern Europe. The date 9750€200 b.p. which is quite imprecise, does not confirm this chronological assignment. The existence of a hiatus in this local zone could explain this discrepancy. LPAZ 3: This zone corresponds to the beginning of the Holocene. Betula achieves maximum values, Ulmus is the first mesophilous tree which expands, Picea shows a continuous curve, all of the herbaceous taxa decrease and Pinus experiences a new expansion. At that time the vegetation around the site probably consisted of a mosaic of open Pinus and Picea stands mixed with Betula, whereas Ulmus stands were developing in warmer places. The date of 9150€350 b.p., although

lacking in precision, taken with Ulmus pollen records of 15% directly at the base of zone M1.3, indicates that the beginning of the Holocene is probably affected by a hiatus. LPAZ 4 and 5: These zones are characterised by a decline in Pinus and Betula, these pioneer trees being replaced by Picea and mesophilous deciduous trees. Fraxinus, Quercus, and Tilia appear (zone M2.4) and develop into a Fraxinus and Ulmus forest formation (zone 5) in which the role of Quercus is modest. Corylus pollen is not abundant in the spectra. Two dates, 8930€250 b.p. and 8740€160 b.p., enable these zones to be assigned to the Boreal. LPAZ 6: The major feature in this zone is the absolute maximum of Fraxinus associated with Ulmus. This event reflects the major expansion of the Ulmus and Fraxinus forest, in which Acer, Hedera, and Viscum are also regularly recorded. Corylus frequencies maintain themselves at a first plateau of 15–20%. The date 7840€230 b.p. places this zone at the beginning of the Atlantic. LPAZ 7: This zone is present only in M2.7. Corylus percentages reach a second plateau at about 25%, whereas the Fraxinus optimum has passed. LPAZ 8: Zone M1.8 corresponds to 25 cm of a peat very rich in humus, with abundant fragments of Eriophorum fibres. It is marked by a sudden increase in Carpinus percentages (10% and 20% in the two first spectra respectively), an increase by a half in Corylus and a strong decline in Ulmus and Fraxinus. These extraordinary peculiarities, of which no equivalent is found in sequence M2, directly suggest an anomaly due to an accident in coring. The series of dates obtained for two levels of zone M1.8, in spite of an inversion, confirms the accidental origin of this disturbance. Water circulation in the peat through small channels related to the above mentioned water holes might explain this. LPAZ 9: In this zone, Carpinus pollen disappears from the spectra while Corylus percentages maintain themselves at the second plateau they had reached in zone 7. The pollen content of this zone is very similar to that of zone 7. The disturbance recorded in zone 8 does not seem to have affected the forest dynamics. A peak in Ericaceae (Empetrum type) recorded at the centre of the peat bog may reflect slightly drier conditions in the bog.

118

Fig. 3 Mohos 1 pollen diagram of relative frequencies (SA Subatlantic, SB subboreal, AT Atlantic, BO boreal, LG Late Glacial)

119

120

Fig. 4 Mohos 2 relative pollen frequency diagram of (SA Subatlantic, SB subboreal, AT Atlantic, BO boreal)

121

122

LPAZ 10: The Corylus maximum and the reappearance of a few occurrences of Carpinus pollen are the salient features of this zone. LPAZ 11: In this zone, a strong decrease in Corylus mostly favours Picea (M2.11). In the sequence M1 a few occurrences of cereal are associated with a decrease in the mixed oaks. It indicates that human activities were affecting the deciduous formations. The effects of this event lasted until about 6230€240 B.P. LPAZ 12, 13, 14: These correspond to a period marked by a steady increase in Carpinus and a decrease in Corylus. At the beginning of zone M1.12, i.e. at about 6230€240 b.p., Corylus percentages attain a maximum while Carpinus is absent. Between the beginning and the end of zone M1.12, Corylus values decrease by about 50%. Zone M2.13 is a unique zone that has no equivalent in the sequence M1. It is characterised by two spectra in which Fagus values exceed 40%. It seems likely that this anomaly has the same accidental origin as that observed in zone M1.8. LPAZ 15: This zone corresponds to a maximum expansion of Carpinus the beginning of which is dated to about 4400 B.P. in the two sequences (4550€100 B.P. in M2 and 4220€180 b.p. in M1). At that time vegetation around the crater was a Carpinus forest with Picea covering less sunny places on the summits. The end of this Carpinus optimum expansion is dated between 3310€170 b.p. and 2910€90 b.p. LPAZ 16: In this zone, which is present only in M2, Fagus replaces Carpinus. Unfortunately a sample derived from sequence M2 at 420 cm could not be dated by routine methods as it proved to have a too low carbon content. LPAZ 17: This zone is marked by the absolute dominance of Fagus, with values only exceptionally below 50%. At that time, Fagus extended over the whole periphery of the Mohos¸ crater, gradually supplanting Picea, which could probably only maintain itself on the summits. Abies, represented by sporadic pollen occurrences so far, plays a modest role during this Fagus optimum that lasts at least 20 centuries, from 2910€90 b.p. until after 900€150 B.P. Herbs are not abundant in this zone, but human activity is attested in M1 by regular records of cereal pollen (Secale) and Plantago lanceolata.

LPAZ 18: This zone belongs to historical times. Its base is dated to 540€160 B.P. in sequence M1.18. Juglans, cereals and Secale are regularly recorded. Fagus begins to decline at the end of the zone. The modest increase in Pinus pollen percentages is due to an anthropogenic opening of forest environments, favouring this opportunistic pioneer everywhere in the region. LPAZ 19: In this modern zone evidence is found of a peak in agricultural activities. Zea pollen is present in all the pollen spectra recorded from M1.19. Fagus percentages fall abruptly to values below 20%.

Discussion and conclusion At Mohos¸ the peat layer directly overlies the tephra ejected by the nearby Sfnta Ana crater (Juvign et al. 1994). The date proposed for this event is ca. 10,700 b.p. (ibid.). The pollen analysis of sequence M1 confirms these tephro-chronological data, in that in this sequence sedimentation begins at the end of the Late Glacial with a Cyperaceae peat (Table 1) covering the three first zones of the profile. In M1.2, the Younger Dryas cooling is manifested by a decrease in Pinus and an extension of herbaceous steppe taxa, notably Artemisia. These conditions are fairly similar to those described at the site of Taul Zanogutii (Farcas et al. 1999) in the south-western part of the Romanian Carpathians (Fig. 1). The first periods of the Holocene (Preboreal) are probably absent, even in sequence M1. Although lacking in precision, the date 9150€350 b.p. indicates the Preboreal/Boreal transition, and the abrupt change in the pollen values of Betula, Ulmus, Pinus, and Cyperaceae at the limit of zones M1.2 and M1.3 suggests a hiatus that may cover several centuries. At this middle altitude site, the forest dynamics of mesothermophilous deciduous trees are clearly recorded. It is one of the regions in Europe where Ulmus played a pioneer role before 9150€350 b.p., competing with Betula. The date obtained for the bottom of zone M2.5 indicates that Fraxinus, Quercus, Tilia, and Corylus became established almost simultaneously around 8740€160 B.P. However, it seems that a slightly greater age can be postulated, considering that a more advanced stage of these forest dynamics is dated to 8930€250 b.p. in the middle of zone M1.4. Moreover, these ages seem rather young compared with the date of 9505€85 b.p. obtained at Taul Zanogutii (Fig. 1) for a similar forest stage (Farcas et al. 1999). Ulmus, Fraxinus, and Tilia showed a long simultaneous optimum expansion that lasted until about 3000 B.P. These three trees were the main components of the mountain mixed deciduous forest, with Quercus frequencies rarely exceeding 10%.

123

Fig. 5 Location map of the cited archaeological sites

Corylus achieved a major expansion from 6660€250 to 5070€220 b.p. In the two sequences, pollen zone 11 (Table 2) is a good synchronous reference. Prior to this zone, Carpinus pollen is recorded more or less regularly and always with values below 0.5% from the top of zone 9, that is to say 6660€250 b.p. (Table 2). Between zone 11 and zone 15, the Carpinus pollen curve differs markedly in its aspect from one sequence to the other. In M1 Carpinus frequencies remain below 1% in all the spectra, whereas in M2 the Carpinus curve is continuous, with values nearly always exceeding 5%. The age inferred for the local establishment of Carpinus would have been quite different if only one sediment sequence had been available. The bottom of zone 12, corresponding to the local establishment of Carpinus, yielded an age of 6230€240 b.p. in M1 and 5780€90 b.p. in M2. Therefore an age of about 6000 B.P. can reasonably be accepted for this event. The beginning of the Carpinus maximum expansion is dated 4550€100 b.p. in M2, and ends at about 3000 b.p. with the Fagus expansion. Because of its poor dispersal capacity, Fagus pollen is a good indicator of the local presence of beech forest, even when its frequencies are low (Heim 1970; Reille 1975; de Beaulieu 1977). Fagus pollen is regularly recorded in the spectra right from the bottom of zone 6, for which an estimated age of 8000 B.P. can be proposed. These results clearly suggest that Fagus has been present near the site since this date. Its local expansion, which led to its absolute dominance, took place only at about 4000 b.p. Beech pollen was recorded in some other diagrams from the Romanian Carpathians (Taul Zanogutii, Steregoiu, Ic Ponor) starting at 7500–8000 b.p. (Farcas et al. 1999; Bj rkman et al. 2002; Bodnariuc et al. 2002) while in Semenic Mountains it was dated to 9500 b.p. (R sch and Fischer 2000). The first indications of human activities are observed in sequence M1, between 6660€250 and 6230€240 b.p. with the first occurrences of cereal pollen. Zone 11 is

probably the result of anthropogenic disturbance. In the pollen spectra cereals are recorded during the Carpinus optimum and throughout the Fagus phase, and become more regular with the appearance of Plantago lanceolata (zones 17, 18, 19). As in many other regions in Europe, human activities were one of the factors that favoured Fagus expansion and propagation (K ster 1997). Archaeological evidence to confirm pollen data was found in some sites near the Mohos crater (Fig. 5). Traces of the Boian culture (middle Neolithic) dated to 5500–6000 b.p. were found at Bra˘dut¸ (altitude 690 m) and Turia (Szkely 1998). Some radiocarbon dates were obtained at Malnas¸Ba˘i (5300–5600 B.P.) for pottery of the Cucuteni culture (Laszlo 1997). Albis¸, Peteni and Zoltan settlements were confined to the Bronze Age (3500–4000 b.p.) (Szkely 1980). The first occurrences of Juglans pollen are recorded rather late, at 540€160 B.P. Thereafter, regional disturbances manifest themselves through the decrease of Fagus percentages in M2.18 (or their fluctuations in M1.18) and a notable increase in Pinus, this opportunistic pioneer taking advantage of an opening in forest environments. The last zone (zone 19), in which Zea pollen is recorded, belongs to modern times. It is the only zone where the AP is constantly below 80%. In terms of biogeography, Romania and the Carpathian mountains constitute a key territory in which the fauna and the flora from the eastern European plains, Central Europe and the Mediterranean part of the Balkan Peninsula are brought into contact with each other. It is a known fact that the southern part of the Carpathian Mountains was a refuge area for temperate and mountain forest taxa during the last glacial (Huntley and Birks 1983). During the Holocene these regions were probably a major tree route for the re-colonisation of central Europe. Quite recent work on the genetics of deciduous oak populations (Kremer et al. 2000) has revealed the existence in Romania of very particular “haplotypes”. These have resulted either from the existence of older indigenous relict populations or from the crossing of populations originating from numerous refuges (eight species of deciduous oaks have been described in Romania) (Tutin et al. 1964–1980). The present paper contributes to a programme of systematic dating of key pollen sequences aimed at regional differentiation of vegetation dynamics and the location of refuges. The beginning of the Mohos¸ sequence is too young to enable us to identify possible refuges. The regular occurrences of Alnus, Ulmus, Quercus, and Picea at the base of the Mohos¸ 1 sequence during the Younger Dryas merely suggest a regional immigration, at least during the Late Glacial interstadial. The other taxa, either mesophilous taxa or mountain taxa, arrived later. There is a certain resemblance between the Holocene dynamics of Mohos¸ and the evidence from western Europe during the last interglacial (Zagwijn 1996): Ulmus was the first mesophilous deciduous tree to become established and the Corylus expansion occurred after the major expansion of

124

the mixed oak-forest (in which Quercus was not the dominant tree). In Romania the Corylus optimum is not correlated with the Boreal as is classically occurs in western Europe, but rather with the Atlantic chronozone (Diaconeasa and Farcas 1996) which corresponds to a climatic optimum attested by the presence of abundant Hedera and Viscum, both taxa being indicative of mild winters. If one takes into account the dates at which the trees in the mountain stage appeared and expanded, it appears that the regional presence of Fagus is recorded quite early at Mohos¸, three millennia earlier than in the Ca˘limani mountains (Farcas et al. 1999). This suggests that this expansion took place from populations situated to the south and south-west part of the mountain range. There are occurrences of beech pollen in some Late Glacial diagrams from Sucho Ezero, Kupena (Bozilova et al. 1996) and Sedmo Rilsko (Bozilova and Tonkov 2000) in Bulgaria, Ljubljana (Culiberg and Sercely 1996) in Slovenia, Ioannina (Bottema 1974) and Kopa s (Allen 1986) in Greece and from Lake Malik (Denfle et al. 2000) in Albania. As far as the Holocene is concerned in Bulgaria, near the frontier with Romania, at the coastal lake of Durankulak, a continuous Fagus curve is recorded from 6000 b.p. (Bozilova et al. 1996). However the expansion of forests dominated by Fagus in Romania is dated to around 4000 b.p. The same applies to south-eastern Poland (Ralska-Jasiewiczowa and Latalowa 1996). In the Romanian Carpathians the major Carpinus expansion is dated to between 4000 b.p. at Steregoiu (Feurdean et al. 2001) and 6500 b.p. at Taul Zanogutii (Farcas et al. 1999). The Holocene vegetation dynamics at the different vegetation levels of the Carpathian mountains reveal several original features, but the first 14C dates obtained do not enable us to conclude that there were any marked regional differences. However, the available data are as yet insufficient and further efforts will have to be made to acquire new 14C-dated sequences. Acknowledgements The authors thank Florin Farcas for the help during the fieldwork, G. Digerfeldt for the lithological description of the cores and Anne Vaillant for the sample preparation. We would like also to thank H.J. Beug and an anonymous reviewer for the critical and constructive comments on the first version of the manuscript.

References Allen H (1986) Late Quaternary of the Kopais Basin, Greece: sedimentary and environmental history. Thesis, Cambridge University Beaulieu JL de (1977) Contribution pollenanalytique  l’histoire tardiglaciaire et holocne de la vgtation des Alpes mridionales fran aises. Thse s sciences, Aix-Marseille3 Bj rkman S, Feurdean A, Cinthio K, Wohlfarth B, Possnert G (2002) Late glacial and early Holocene vegetation development in the Gut iului Mountains, northwestern Romania. Quat Sci Rev 21:1039–1059

Bodnariuc A, Bouchette A, Dedoubat JJ, Otto T, Fontugne M, Jalut G (2002) Holocene vegetational history of the Apuseni Mountains, central Romania. Quat Sci Rev 21:1465–1488 Bottema S (1974) Late Quaternary vegetation history of northwestern Greece. Thesis, Rijksuniversitt, Groningen Bozilova E, Filipova M, Filipovch L, Tonkov S (1996) Bulgaria. In: Berglund BE, Birks HJB, Ralska-Jasiewiczova M, Wright HE (eds) Palaeoecological events during the last 15000 years. Regional syntheses of palaeoecological studies of lakes and mires in Europe, pp 701–728 Bozilova E, Tonkov SB (2000) Pollen from Lake Sedmo Rilsko reveals southeast European postglacial vegetation in the highest mountain area of the Balkans. New Phytol 148:315–325 Coldea Gh (1991) Prodrome des associations vgtales des Carpates du sud-est (Carpates roumaines), Documents Phytosociologiques, Camerino 13:539 Culiberg M, ercelj A (1996) Slovenia. In: Berglund BE, Birks HJB, Ralska-Jasiewiczova M, Wright HE (eds) Palaeoecological events during the last 15,000 years. Regional syntheses of palaeoecological studies of lakes and mires in Europe, pp 687– 700 Denfle M, Lzine AM, Fouache E, Dufaure JJ (2000) A 12,000year pollen record from Lake Maliq, Albania. Quat Res 54:423–432 Diaconeasa B, Farcas S (1996) Mixed oak forest; their Lateglacial and Holocene evolution and dynamics in Romania (in Romanian with English summary and French rsum). Contrib Botanicae 1995–1996, Cluj-Napoca, pp 103–115 Farcas S, Beaulieu JL de, Reille M, Coldea G, Diaconeasa B, Goeury C, Goslar T, Jull T (1999) First 14C datings of late Glacial and Holocene pollen sequences from Romanian Carpathes. C RAcad Sci Paris 322:799–807 Feurdean A, Bj rkman L, Wohlfarth B (2001) A paleoecological reconstruction of the Late Glacial and Holocene based on multidisciplinary studies at S¸teregoiu site (Gut i Mts., NW Romania). Stud Univ “Babes¸-Bolyai” Geol 46(2):125–140 Goeury C (1997) GpalWin: gestion, traitement et reprsentation de la palocologie. In: XV-me Symposium de l’AP.F, Lyon, Sept 1997, 31 pp Goeury C, Beaulieu JL de (1979) A propos de la concentration du pollen  l’aide de la liqueur de Thoulet dans les sdiments minraux. Pollen Spores 21:239–251 Heim J (1970) Les relations entre les spectres polliniques rcents et la vgtation actuelle en Europe occidentale. Editions Derouaux, Lige Huntley B, Birks HJB (1983) An atlas of past and present pollen maps for Europe: 0–13,000 years ago. Cambridge University Press, Cambridge Juvign E, Gewelt M, Gilot E, Hurtgen C, Seghedi I, Szakacs A, Gbris G, Hadnagy A, Horvth E (1994) Une ruption vielle d’environ 10700 ans (14C) dans les Carpates orientales (Roumanie). C RAcad Sci Paris 318:1233–1238 Kremer A, Petit RJ, Csaikl JL, Beaulieu JL de, Bordacs S, Brewer S, Burg K, Cheddadi R, Coart E, Cottrell J, Cundall N, Deans D, Fineschi S, Finkeldey R, Gillies A, Goicoechea P, Gracan J, Grossoni P, Hall CL, Jensen JS, Kleinschmit J, K nig A, Lowe A, Madsen SF, Matyas G, Munro R, Olalde M, Olesdska I, Pemonge MH, Popescu F, Reille M, Slade D, Steinhoff S, Tabbener HE, Van Dam B, De Vries SMG, Ziegenhagen B (2000) Range wide distribution of the chloroplast and pollen deposits in European white oaks: inferences about colonisation routes, impact of chloroplastic divergence on phenotypic differentiation. IUFRO Conference “Oak 2000”, Zagreb, Abstract Volume, 5 pp K ster H (1997) The role of farming in the Postglacial expansion of beech and hornbeam in oak woodlands of central Europe. Holocene 7:239–242 Laszlo A (1997) Malnas Bai, Cronica cercetarilor arheologice 1983–1992 [The chronicle of archaeological researches 1983– 1992]. The National Commission for Archaeology, Bucharest, pp 52–53

125 Pop E (1960) Mlas¸tinile de turba din Republica Populara˘ Romna˘ [Peat bogs from Romania]. Editura Academiei Republici Populare Romne, Bucuresti. Pop E, Diaconeasa B (1967) Analiza palinologica˘ a turbei din tinovul Mohos¸ (Tus¸nad) [Pollen analyses of the peat from Mohos peat bog]. Contribut¸ii botanice, Cluj-Napoca, pp 297– 303 Ralska-Jasiewczowa M, Latalowa M (1996) Poland. In: Berglund BE, Birks HJB, Ralska-Jasiewiczova M, Wright HE (eds) Palaeoecological events during the last 15,000 years. Regional syntheses of palaeoecological studies of lakes and mires in Europe, pp 404–472 Reille M (1975) Contribution pollenanalytique  l’histoire tardiglaciaire et holocne de la vgtation de la montagne corse. Thse s sciences, Marseille 3, 206 pp

R sch M, Fischer E (2000) A radiocarbon dated Holocene pollen profile from the Banat Montains (southwestern Carpathians, Romania). Flora 195:277–286 Szkely Z (1980) Archaeological considerations on the Bronze Age settlement of Peteni, Covasna county. Archaeological Materials and Researches, Tulcea, pp 129–134 Szkely Z (1998) Bradut, Covasna county, in the chronicle of archaeological researches—the 1997 campaign, Calarasi, 20– 24 May 1998, pp 9–10 Tutin TG, Heywood VH, Burger NA, Moore DM, Valentine DH, Walters SM, Webb DA (1964–1980) Flora Europaea, vol 1 (1964); vol 2 (1968); vol 3 (1972); vol 4 (1976); vol 5 (1980). Cambridge University Press, Cambridge Zagwijn WH (1996) An analysis of Eemian climate in Western and Central Europe. Quat Sci Rev 15:451–469