Palaeoenvironment in north-western Romania during the last 15,000 years

Angelica Feurdean

Dedicated to Ovidiu Feurdean

Avhandling i Kvartärgeologi Thesis in Quaternary Geology No. 3

Department of Physical Geography and Quaternary Geology, Stockholm University 2004

1 ISBN

2 Palaeoenvironment in north-western Romania during the last 15,000 years by

Angelica Feurdean

Department of Physical Geography and Quaternary Geology, Stockholm University, SE-106 91 Stockholm

This thesis is based on work carried out as Ph.D. stu- Appendix III: Feurdean A. (in press). Holocene forest dent at the Department of Paleontology, Faculty of Biol- dynamics in north-western Romania. The Holocene. ogy and Geology, Babes-Bolyai University, Cluj-Napoca, Romania between Oct. 1998 and Sept. 2002 and later as Appendix IV: Feurdean A. & Bennike O. Late Quater- Ph.D. student in Quaternary Geology at the Department nary palaeocological and paleoclimatological reconstruc- of Physical Geography and Quaternary Geology, tion in the Gutaiului Mountains, NW Romania. Manu- Stockholm University 2002-2004. The thesis consists of script submitted to Journal of Quaternary Science. four papers and a synthesis. The four papers are listed below and presented in Appendices I-IV. Two of the pa- Fieldwork in Romania has been jointly performed with pers have been published (I, II), one is in press (III) and Barbara Wohlfarth and Leif Björkman (Preluca Tiganului, the fourth has been submitted (IV). The thesis summary Steregoiu, Izvoare) and with Bogdan Onac (Creasta presents an account of earlier pollenstratigraphic work Cocosului). I am responsible for the lithostratigraphic done in Romania and the discussion focuses on tree description of all sediment and peat cores, for sub-sam- dynamics during the Lateglacial and Holocene, based pling and laboratory preparation. I have done all mineral on recent results from Romania. magnetic measurements and loss-on-ignition analysis (except for the lowermost part of Preluca Tiganului, which was the topic of an examination paper by Kajsa Cinthio). Appendix I: Björkman L, Feurdean A, Cinthio K, I have analyzed all pollen samples from Preluca Tiganului Wohlfarth B, & Possnert G. 2002. Lateglacial and early and Steregoiu, except for the uppermost part of Preluca Holocene vegetation development in the Gutaiului Moun- Tiganului and Steregoiu, which was done by Leif tains, NW Romania. Quaternary Science Reviews 21: Björkman. The macrofossil analyses were performed by 1039-1059. Ole Bennike. As co-author of paper I and II I have been responsible for data collection, analyses, interpretation Appendix II: Björkman L, Feurdean A, & Wohlfarth B. and have contributed to the text. As first author in paper 2003. Lateglacial and Holocene forest dynamics at IV, I have been responsible for data collection, analyses Steregoiu in the Gutaiului Mountains, NW Romania. of the pollen samples, data interpretation, illustrations Review of Palaeobotany and Palynology 124: 79-111. and text.

3 Contents

ABSTRACT...... 6

INTRODUCTION...... 7

BACKGROUND...... 10 Earlier pollenstratigraphic investigations in Romania...... 10 The vegetation succession in the Gutaiului Mountains...... 11 Ongoing palaeoenvironmental and palaeoclimatic studies in Romania...... 11 Objectives of the doctoral thesis...... 11 . STUDY AREA...... 12 Climate...... 1 3 Geology ...... 13 Vegetation...... 13

MATERIAL AND METHODS...... 15 Fieldwork and site selection strategy...... 15 Laboratory work...... 16 Pollen and charcoal analysis...... 16 Macrofossil analysis...... 16 Mineral magnetic analysis...... 17 Loss on ignition (LOI)...... 17 Radiocarbon dating...... 17

RESULTS...... 17 Summary of papers I-IV...... 17 Paper I...... 17 Paper II...... 18 Paper III...... 18 Paper IV...... 19

Unpublished results...... 20 Fire history...... 20 Macrofossil results...... 21

DISCUSSION ...... 22 Glacial refugia...... 22 Forest development in the study area during the last ca. 18,000 cal. yr BP and comparison to other sites in Romania ...... 22 Last Glacial Maximum (LGM) ...... 22 Lateglacial forest development...... 24 Holocene forest development...... 27 Human indicators in the study area...... 33 Climatic fluctuations during the Lateglacial and Holocene as reflected by changes in the vegetation...... 34 >14,700 cal. yr BP...... 34 Lateglacial (14,700 – 11,500 cal. yr BP)...... 34 Holocene...... 37

CONCLUSIONS...... 38

ACKNOWLEDGMENTS...... 39

4 SVENSK SAMMANFATTNING...... 39

REZUMAT IN LIMBA ROMANA...... 40

REFERENCES...... 41

APPENDICES...... 47 I: Björkman L, Feurdean A, Cinthio K, Wohlfarth B, & Possnert G. 2002. Lateglacial and early Holocene vegetation development in the Gutaiului Mountains, NW Romania. Quater- nary Science Reviews 21: 1039-1059

II: Björkman L, Feurdean A, & Wohlfarth B. 2003. Lateglacial and Holocene forest dynamics at Steregoiu in the Gutaiului Mountains, NW Romania. Review of Palaeobotany and Palynology 124: 79-111

III: Feurdean A. (in press). Holocene forest dynamics in north-western Romania. The Holocene

IV: Feurdean A. & Bennike O. Late Quaternary palaeocological and paleoclimatological reconstruction in the Gutaiului Mountains, NW Romania. Manuscript submitted to Journal of Quaternary Science

5 Palaeoenvironment in north-western Romania during the last 15,000 years

Angelica Feurdean

ABSTRACT

The objectives of this thesis are to establish a chronological framework for environmental changes during the last 15,000 years in northwest Romania, to reconstruct the vegetation development, and to evaluate the underlying processes for forest dynamics. Furthermore, an overview of earlier and ongoing pollenstratigraphic work in Romania is provided. Sediments from two former crater lakes, Preluca Tiganului and Steregoiu, situated in the Gutaiului Mountains, on the western extremity of the Eastern Carpathians at 730 m and 790 m a.s.l., respectively were obtained and analysed for high- resolution pollen, macrofossils, charcoal, mineral magnetic parameters and organic matter. The chronostratigraphic framework was provided by dense AMS 14C measurements. Cold and dry climatic conditions are indicated by the occurrence of open vegetation with shrubs and herbs, and cold lake water prior to 14,700 cal. yr BP. The climatic improvement at the beginning of the Lateglacial interstadial (around 14,700 cal. yr BP) is seen by the development of open forests. These were dominated by Pinus and Betula, but contained also new arriving tree taxa, such as Populus, Alnus and Prunus. The gradual establishment of forests may have led to a stabilization of the soils in the catchment. Between ca. 14,100 and 13,800 cal. yr BP the forest density became reduced to stands of Pinus, Betula, Alnus, Larix and Populus trees and grassland expanded, suggesting colder climatic conditions. Picea arrived as a new taxon at around 13,800 cal. yr BP, and between 13,800 and 12,900 cal. yr BP, the surroundings of the sites were predominantly covered by Picea forest. This forest included Betula, Pinus, Alnus, Larix and Populus and, from 13,200 cal. yr BP onwards also Ulmus. At ca. 12,900 cal. yr BP, the forest became significantly reduced and at 12,600 cal. yr BP, a recurrence of open vegetation with stands of Larix, Pinus, Betula, Salix and Alnus is documented, lasting until 11,500 cal. yr BP. This distinct change in vegetation may by taken as a strong decline in temperature and moisture availability. At the transition to the Holocene, at ca. 11,500 cal. yr BP, Pinus, Betula and Larix quickly expanded (from small local stands) and formed open forests, probably as a response to warmer and more humid climatic conditions. At 11,250 cal. yr BP Ulmus and Picea expanded and the landscape became completely forested. The rapid increase of Ulmus and Picea after 11,500 cal. yr BP may suggest the existence of small residual populations close to the study sites during the preceding cold interval. Ulmus was the first and most prominent deciduous taxa in the early Holocene in the Gutaiului Mountains. From ca. 10,750 cal. yr BP onwards Quercus, Tilia, Fraxinus and Acer expanded and Corylus arrived. A highly diverse, predominantly deciduous forest with Ulmus, Quercus, Tilia, Fraxinus, Acer, Corylus and Picea devel- oped between 10,700 and 8200 cal. yr BP, which possibly signifies more continental climatic conditions. The develop- ment of a Picea-Corylus dominated forest between 8200 and 5700 cal. yr BP is likely connected to a more humid and cooler climate. The establishment of Carpinus and Fagus was dated to 5750 cal. yr BP and 5200 cal. yr BP, respectively. The dominance of Fagus during the late Holocene, from 4000 cal. yr BP onwards, may have been related to cooler and more humid climatic conditions. First signs of human activities are recorded around 2300 cal. yr BP, but only during the last 300 years did local human impact become significant. The vegetation development recorded in the Gutaiului Mountains during the Lateglacial is very similar to recon- structions based on lowland sites, whereas higher elevation sites seem not to have always experienced visible vegeta- tion changes. The time of tree arrival and expansion during the past 11,500 cal. yr BP seems to have occurred almost synchronously across Romania. The composition of the forests during the Holocene in the Gutaiului Mountains is consistent with that reconstructed at mid-elevation sites, but differs from the forest composition at higher elevations. Important differences between the Gutaiului Mountains and other studied sites in Romania are a low representation of Carpinus and a late and weak human impact. The available data sets for Romania give evidence for the presence of coniferous and cold-tolerant deciduous trees before 14,700 cal. yr BP. Glacial refugia for Ulmus may have occurred in different parts of Romania, whereas the existence of Quercus, Tilia, Corylus and Fraxinus has not been corroborated.

Keywords: Northwest Romania, Gutaiului Mountains, pollenstratigraphy, macrofossil remains, Lateglacial, Holocene, tree refugia, tree dynamics, past climate, human influence.

Department of Physical Geography and Quaternary Geology, Stockholm University, SE-106 91 Stockholm

6 ward, 1988; Prentice et al. 1993; Sykes et al. 1996). Introduction Many studies have shown a clear spatial response of the vegetation to the distinct climatic variations dur- Paleoenvironmental and paleoclimatic research provides ing the Weichselian Lateglacial in Europe. In some re- an insight into the extent, timing and causes of climatic gions the response was rather strong, e.g. around the and environmental changes, which have occurred in the North Atlantic, while it was less prominent in continen- past. Palaeo-studies have during the past years gained tal areas. These differences may have been due to the increasing attention, because they allow placing the con- spatially varying intensity of temperature/precipitation temporary environmental and climatic changes in a changes, to the physiological tolerance of the species longer time perspective. Important archives for involved, to inter-species competition and migration palaeoenvironmental and palaeoclimatic research are ice speed. Although climatic fluctuations during the Ho- cores, marine and lacustrine sediments, peat bogs, tree locene in general were not as pronounced as compared rings and speleothems. An array of different methods to the glacial stages, new studies have shown that even can be applied to each of these archives to obtain as the early Holocene climate was relatively unstable and much climatic and environmental information as possi- characterized by several oscillations (e.g. Bond et al. ble. 1997, 2001), informally termed the “Preboreal Oscillation” Pollen and plant macrofossil analyses are for terres- at ca. 11,300 cal yr BP (Behre, 1966; Björck et al. 1997), trial archives, such as lake sediments and peat deposits, the “10,300 cal yr BP” event (Björck et al. 2001) and the the most common tools to reconstruct past vegetation so-called “8200 cal yr BP” event (Bond et al. 1997). These and biogeography, past climate and human impact on fluctuations were initially only recorded around the North the vegetation. Macrofossil analysis has been regularly Atlantic region and on Greenland, but also appear to employed since the 19th century, while pollen analysis have occurred in continental areas, where they seem to was developed during the early part of the 20th century have been relatively pronounced (von Grafenstein et al. (Reid 1899; Erdtman, 1934; Iversen 1941, 1949; von Post 1998, 1999; Ralska-Jasiewiczowa et al. 1998; Tinner and 1916, 1946; Watts 1959; West 1977; Punt and Clarke Lotter, 2001; Magny et al. 2003; Wick et al. 2003). It has 1984; Birks 1986, 2001; Fægri and Iversen 1989; Reille been shown that these short-term climatic variations must 1992, 1995; Bennett and Willis 2001; Birks 2001). Al- have been severe enough to have an impact on the exist- though these two techniques have been used and de- ing vegetation (Ralska-Jasiewiczowa et al. 1998; Ammann veloped in parallel, pollen analysis is still the major tool and Birks, 2000; Tinner and Lotter, 2001; Wick et al. 2003). and probably the most widely applied method to inves- During the past, palaeoenvironmental research has tigate past vegetation dynamics. This is due to the large mainly focused on regions around the North Atlantic production of pollen and their dispersion capability, which and in central Europe and only few records have been leads to pollen grains being relatively easy deposited in available for southeast Europe. Pollen stratigraphy has lakes and on peat. Pollen analysis has, however, some generally been the main tool to infer changes in vegeta- drawbacks (see Objective chapter of the thesis summary) tion composition for southeast Europe, but most of the and palaeoecologists have tried to develop and improve pollen records have low stratigraphic resolution and poor the method over time. Andersen (1974) made the first chronological frameworks. Attempts to reconstruct past attempts by using the so-called “correction factor”. This vegetation changes and to locate refugia during cold approach was followed by numerical modelling to pre- glacial stages had therefore often to rely on such low- dict catchment scale and size of the sedimentary basins resolution studies (Huntley and Birks, 1983; Bennett et (Prentice 1988; Sugita 1994, 1999). More recently models al. 1991; Willis 1994). West (1977) and Huntley and Birks of pollen dispersal have been developed based on a com- (1983) were among the first to compare pollen bination of pollen traps, vegetation surveys, climate stratigraphies across Europe to decipher the immigra- records and fossil pollen assemblages to evaluate the tion routes for tree taxa following the last glacial stage. relationship between pollen assemblages and the com- Their assumption was that the locality where a tree taxaon position of the vegetation (Hicks 1994, 2001; Broström appeared for the first time during the postglacial would 2002). This knowledge combined with high-resolution, correspond to the place where the taxaon had its glacial chronologically well-constrained pollen stratigraphy has refugia. Huntley and Birks (1983) could show that glacial increased the reliability of pollen analysis and allows a refugias must have been concentrated to southern Eu- more objective interpretation of vegetation characteris- rope and may have existed in regions such as the Bal- tics through time. In addition it is necessary to consider kans, the Alps, the Carpathians and the Italian moun- the present distribution of trees across Europe, which is tains. Tree species survived in these areas during the limited by the following climatic parameters: the number cold periods of the Quaternary and spread northwards of growing-degree days controls their northern limit, the at the beginning of an interglacial. Later, Bennett et al. temperature of the coldest month constrains their east- (1991) suggested that other factors also need to be con- ern limit and water availability their southern limit (Wood- sidered, such as e.g. the modern distribution of tree spe-

7 cies in Europe; tree dynamics during the last cold stage Central Southern and during intervening interstadials; climatic conditions Romania Europe Scandinavia during glacial stages and the physiography of southern Forest phases Firbas Blytt ( 1881) Europe. However, for a number of areas, these discus- according to zonation Sernander (1890) sions had to be based on old pollen stratigraphic stud- Pop (1942) system Nilsson (1961) (1949,1952) ies and it became evident that multidisciplinary ap- proaches with high temporal, spatial and ecological reso- Picea-Fagus-Abies X lution were needed to address precise environmental Sub-Atlantic Spruce- beech- fir IX reconstructions. Recently published palaeoenvironmental records Picea-Carpinus VIII Sub-Boreal from south-eastern Europe now allow more accurate es- Spruce-hornbeam VIIb timations of the palaeovegetation during periods of dis- VIIa tinct climatic changes. (Willis 1992a, b; Willis et al. 1995, Picea-QM-Corylus Atlantic 1997, 2000; Denèfle et al. 2000; Rudner and Sümegi, 2001; Spuce- mixed oak VI Sümegi and Rudner, 2001; Tonkov et al. 2002; Stefanova hazel Vb Late Boreal and Ammann 2003). They also show that refugia for de- Pinus-Picea Boreal ciduous trees even existed in regions situated north of Pine - spruce Va the Balkan Peninsula (e.g. Hungary). Despite the emer- IV Preboreal

gence of these new records, Romania has largely re- e s

a III Younger Dryas mained a “white spot” on the palaeoenvironmental map h p

of Europe. Data sets from this part of Europe are, how- e Alleröd

n II i

P Some spruce, ever, important to assess the spatial variability of past

birch, alder Ic Older Dryas s

changes in vegetation and climate and to reconstruct u Picea-Betula-Alnus n i Ib Bölling tree migration routes at the begginig of the present inter- P glacial. Only by integrating this region into a European Ia Oldest Dryas context will it be possible to obtain a complete picture of the palaeoenvironmental development. This thesis thus focuses on a reconstruction of the Lateglacial and Holocene vegetation development in Figure 1. Pop’s (1942) Lateglacial and Holocene forest phases north-western Romania. This has been done through and their correlation to the pollenstratigraphic schemes of high-resolution pollen analyses of two lake/peat bog Firbas (1949, 1952), Blytt (1881), Sernander (1890) and sequences in order to get a good temporal and spatial Nilsson (1961). resolution of Lateglacial and Holocene vegetation changes. Pollen analysis has been the main tool, which was complemented by macrofossil, organic matter, min- eral magnetic analyses and AMS 14C measurements.

s 1881 48 SLOVAKIA 22 24 26 28 UKRAINE 48 s 2061 MOLDOVA 3 48 1 2 5 23 4 s Figure 2. Topographic map of a 6 Tisz Romania showing the location of HUNGARY 20 21 7 earlier studied lake and peat bog 18 Cluj sites. Numbers 1-23 refer to Ta- 17 19 22 46 s 8 UKRAINE ble 1 where more detailed site information is given. 9 46 12 N s 14 13 s s s Black Sea 15 > 2000 m a.s.l. 16 Bukarest Danube 1000–2000 44 11 Danube 400–1000 YUGOSLAVIA 10 200–400 < 200 BULGARIA 0 50 km 44 22 24 26 28 30

8

m

) ) )

7 m 3

) 4 ; 6

o 9 7 6 l ) 9 9 a 6 9 H 7 1 e 1 - 9 1 ( 5 r t (

( 1

,

s 9 o ( i z a a

1 b s u s

u ) ) )

a ( b )

) a a s

7 9 9 B G u b c e u ) e

p 7 6 7 ) ) S 5 5 n n 1 n d d u 9 9 9

0 8 a 6 6 o 7 o

n n 1 1 1 L = ) ) 6 4 c b c

9 9 9

( ( ( a a

0 0 a a

9 9

o d 1 1 1 s i i B i ) ) ) ) ) ) a a ) a ) a ) a ) ) ( ( ( 8 8 1 1 n e

s s s s s 2 8 9 2 8 7 2 2 ( ( 7 8 8 S . . . 9 9 D C D a

c

a a a a a l l l

3 5 2 3 5 5 3 3 5 5 5 ; 1 1 e e e u e u e n a a a d d d u ( (

9 c 9 9 9 9 9 9 9 9 9 9

i n n n n n n n i e ) n n n t t t 1 1 1 1 1 1 1 1 1 1 1 t a a a a a r o o o o o 1 ( e ( ( ( ( ( e a a ( e ( ( a ( ( s s n

c

e c c c c b b c 8 s f a p p p p p p p p p p p p p p p p p p p a a a a a o o l 9 e i i i i i i i o u u o t o o o o o o o o o o o o o o o o 1 P L P P P L P P P P P R D P D P P D P D C C B ( P D P P P A

=

,

s A

u M ; l m Q a l e d r U e o r r

e B o t

f d t

= a n s

c a n

B s i

s ; a s s e d l r i u u s a n g l l b

u e a y y

r e A g r r s l

x o , a i o o u g l b n F n C C e M a i

i

r r S s p g d d Q

P o

, a

;

n n L ,

f s

x a a x x C =

i

u g

i

i n x

l w l

l s s i i s n d b

L l s a l o a a u u

e s n l P e a

S

S S

A n e g i a

m m

s ,

i ; S i

, , ,

l l

,

a b s e u s l b a a a a a a a a a a u l B A n n U U l l l l l u l l l l b

A i l n e u

u u u u u u u u r r u i

n d

t y p t t t t t t t t c t i m

d s o o s r r p e n e

e e e e e e e e o f f e n u s u s

o a a a l B s B B B B B

B B B a B

s s n n

u

u

o

s s , C C C , i , , i , , l , , , , e e s s

n a u a a p y a i a p H r r r a a a a a a

u u u u

r r e l l r e g e e e e e e e e o p o o n g g c = a a c a o f c f f c r a c c a c c c c

e i

i i i i i a a i i i i v v l a s s s C C F

C l P P P P

H P P F F P P P

P e e e

h h

C

r r r

o r r ; r

r r r r r r r r r r u u u g g o o o l f o p l l l o i i o o o o o o o o o o f f f o

f a f f a o a f a f

f f f f f f f

f i h h e

s s s

s v v v c s s s s s , , s s s s s s s e s

h e e e e a s s e e e e e t e e e e e e e c e h h h l u u u

u

u u u u u u u l l l u u u u u u u n u l f g g g g l l l l l ) l l l l l l l l i i i g g a a a a a t o e a a a a a a a a a a a a a

h h h t v v v a a n v n

v v v v v

s v v v v v v v

v i a ;

;

;

a

F F c h h h ) ) ) ) ) )

h i L w n t t t t t m t w w w

w f f w w g g w g w w w w w

t i g o i i i n n n n n o n g o o o o o o s o o o i o o o o o l = l

i l l l l l a h a h l h a l l a a a l l

l

m d

h

L

;

e e r

;

; ; ; n ; n ; n ; ; n ; n n ; , ; ; ; ; g o t s , t i i i i i i t t t t t t t t t t t t t t c c e t n d n L t g

n i n n n n n n

n n n n n n n n n n ( n a t

m m m m m m c s

a a a a a a a a a a a a a a a a a L s e n t n o o o o o o

r r a u n n n n n n n n n n n n n n n i u a i x i i i i i i i i i i i i i i a a d d d d d d r n c m

i g e n e e r a m

e t m i m m m m a a m m m a m m a m a a m m m m l a

p p e s o i i i i i i h o o o o o o o o o o o o o o o l l l l l e l e p p u u m F c d

i i i i i i d

d d d d d d d d d d d d d d h a a n o ,

e

s t T T Q T T T T i n s s s s s s s s s s s s s s

d ( ( ( ( ( ( ( a a p i y y u p

u u u u u u u u u u u u u u n l l e e p a r g i r r n n n n n n c n c n n n n n n n

e M M M M M M M M a a l i a a i i i i i i i i i i i i i i t d M F A P Q E E

P P P P Q P Q P Q C P P Q P Q P Q S Q P P P P e s s u c s i d

m

d a d n

A r r n

e B l a g o e

l

-

t c a

o g g g g g g g g a g g g g B 2 b b b b b i e

S L P B r d P L L L P P B B B B B L A L P L L P A L L L L L e r u g i F

n i s

t t t t t n n a a a a e e e e e w p p p p o

m

,

,

,

,

h i t t t t t t t t t t t t t t t t t t t t t t y y y y y y s a d a a a a a a a a a a a a a a a a a a a a a

a a a a a a e e e e s e e e e e e e e e e e e e e e e e e e l l l l l l p p d p P p S p c p p p p p p p p p c c c p p p p p c c p p r o c e 0 r

4 e c 9 i

d 1 h

-

u m p t

0 4 0 0 0 0 0

i a ) 0 0 5 0 0 0 0 6 4 4 0 0 5 0 0 0 6 0 0 5 0

0 0 t r l 2 0 4 0 3 0 1

2 0 0 0 1 7 4 5 8 0 5 9 4 4 3 0 0 1 g m i 5 9 1 7 7 ( 9

A 5 9 1 1 2 9 1 1 1 9 1 7

3 2 7 3 6 5 9 9 t a r t s

i n n e e i l a l l i e e o u r l h

e p

i . i

t u e g d i i i a r e t h n a

t m t b . e n r u a i

i i n u V s i

i l a

t s g g t u i H

r c i i w

o i r x e

u o s

c H o t i u

u u a a T g M s i a n l u C i i T

t c

m

L e

e e

l r

a m

i t i a

i l

M

a o

L - t

i a a a a s c a r e

c d a

a r v r i n c r n u B a N

Z g u e e n e o a e n

g

a c z

o h s e m o o o e e u e

a d

n u c l l l e s s d e i h t e e l b n o g r i s f u t b r b r e l r l t g n d e u u u a m s l i t l s l e a o a e t o o a o o i o a i r e a r o i a a a c r e u t i o e t a e t c H P F S T B B D S M T M B B C R P S T S E P B S M V C B B s r i e L u

. Q 1

e = e .

l t i o

0

2

1 3 2 5 4 7 3 6 8 1 9 0 3 3 b M S n 3 a 1 2 5 7 7 6 1 4 8 1 9 1 1 2 1 1 1 2 1 1 2 1 2 2 2

Q T

9 Background in pollen diagrams from different parts of Romania, were interpreted as caused by differences in altitude and lati- Earlier pollenstratigraphic investiga- tude of the studied sites (Pop et al. 1965a, 1965b, 1965c; tions in Romania Pop and Ciobanu 1957; Pop and Diaconeasa 1967; Ciobanu 1958, 1960; Diaconeasa 1969, 1977, 1979; Pollenstratigraphic investigations in Romania were pio- Boscaiu and Lupsa 1967; Lupsa 1972, 1977, 1980; Boscaiu neered by Pop (1929, 1932, 1942, 1957, 1958, 1960, 1971) et al. 1983; Buz 1986, 1999; Diaconeasa and Farcas more than 70 years ago. Pop was also the first to estab- 1995-1996, 1998; Farcas 1995, 1996). lish a scheme for the Lateglacial and Holocene forest The above-mentioned authors concluded that Pinus, development by placing the boundaries between differ- Picea, Betula, Salix, Alnus, Quercus, Ulmus, Tilia, and ent forest phases where distinct changes in the compo- possibly Carpinus survived in Romania during the cold sition of the pollen assemblages occurred. An age as- stages of the last glacial. The establishment of Carpinus signment for the different forest phases was obtained in prior to Fagus is opposite as compared to the central comparison to the schemes established by Blytt- European forest succession and was named as a par- Sernander (1881, 1890), Firbas (1949, 1952), Nilsson (1935, ticular forest phase for eastern Europe. The late appear- 1961) and Iversen (1942). Pop’s forest phases were up to ance and expansion of Abies and Fagus was explained recently used as the framework for the general vegeta- by the absence of refugia for these trees in Romanian tion history in Romania (Fig. 1): (Pop 1942, 1960; Ciobanu 1958). Later, Diaconeasa and 1. Pop’s Pinus - phase covered the Lateglacial and Farcas (1995, 1998), Buz (1999) and Farcas (2001) attrib- Preboreal (IV) (Fig. 1). The Lateglacial period was sub- uted the presence of thermophilous tree-pollen grains divided according to Firbas (1949, 1952) into Ia (Oldest during the Lateglacial to contamination during coring Dryas), Ib (Bølling), Ic (Older Dryas), II (Allerød) and III procedure, to re-deposition, or to long-distance trans- (Younger Dryas). The reconstructed vegetation, which port, and only the southern and south-eastern part of was dominated by Pinus, also contained Picea, Betula, Romania was suggested as a glacial refugia for Ulmus, Alnus and Salix. Low numbers of pollen grains of Quercus, Tilia and Fraxinus. Quercus, Ulmus, Tilia and Corylus were noticed during Although radiocarbon dating was increasingly used the zone correlated to the Allerød. during the 1960s, 1970s and 1980s for determining the 2. The Pinus-Picea - phase was by Pop attributed to the age of pollen-zone boundaries and for establishing a late Preboreal and early Boreal, i.e. to zones IV and V of chronostratigraphic framework for individual sequences, Firbas (1949, 1952) (Fig. 1). This time interval was marked pollenstratigraphic investigations in Romania continued by distinct forest reorganization. Pine forests spread to to be carried out without any radiocarbon dates to sup- higher elevations in the Carpathians, spruce forest ex- port an age assignment (e.g., Diaconeasa 1968, 1969, 1977, panded at moderate elevations, and Quercus, Ulmus, 1979; Lupsa 1972, 1977, 1980; Boscaiu et al. 1983; Buz Tilia and Acer pollen grains appeared in low percent- 1986, 1999;Diaconeasa and Farcas 1995-1996, 1998; ages. Farcas 1995, 1996). Further limitations of these earlier 3. The Picea-Quercetum mixtum – phase, which includes studies were: low sampling resolution; a focus on forest Quercus, Ulmus, Tilia and Corylus, was assigned to the pollen taxa and on the Holocene vegetation develop- Boreal and Atlantic, i.e. to zones V, VI and VII of Firbas ment; and, results were often only published in national (1949, 1952) (Fig. 1). Forests of Quercetum mixtum and Romanian journals. Consequently, the Quaternary veg- Picea dominated in the foothill areas, whereas Quercetum etation development in Romania was poorly known out- mixtum co-dominated with Corylus in the lowlands. side the country, which often led to Romania being de- Quercus was abundant within the Quercetum mixtum picted with question marks, when palaeoenviromental communities in northern Romania, while in central and reconstructions for Europe were addressed (e.g., Hunt- southern Romania Tilia was the most important forest ley and Birks 1983, Bennett et al. 1991; Willis 1994; taxa. Renssen and Isarin 2001; Renssen et al. 2001; Ravazzi 4. The Picea-Carpinus – phase was correlated to the 2002). To fill this gap, a selection of the most complete Subboreal, i.e. to zone VIII of Firbas (1949, 1952). The and representative of the earlier investigated sites is Carpinus forest belt occurred between the Picea and shown in Figure 2 and Table 1. the Quercetum mixtum forest belts. Fagus became es- The importance of the earlier pollenstratigraphic work tablished during this time span. in Romania should by no means be underestimated. 5. During the Picea-Fagus-Abies – phase, which Pop However, the hypotheses emerging from these studies attributed to the Subatlantic or zones IX and X of Firbas regarding age assignment of the different forest phases, (1949, 1952), Picea, Fagus and Abies were common trees location of glacial tree refugia and past forest dynamics in hilly and mountain areas, while Quercetum mixtum need to be confirmed by more accurate, high-resolution forest dominated the lowland areas. pollenstratigraphic work and detailed radiocarbon dat- Minor differences in vegetation succession, as seen ing.

10 The vegetation succession in the Ongoing palaeoenvironmental and Gutaiului Mountains palaeoclimatic studies in Romania

In 1942 Pop performed a fairly complete investigation on During the last five years Romania has become the sub- the modern and fossil flora of the Oas-Gutaiului Moun- ject of an increasing number of investigations address- tains (Fig. 3) with the aim to enlarge the botantical and ing the Lateglacial and Holocene environmental devel- palaeobotanical knowledge from northern and northwest- opment. These investigations comprise pollen- and ra- ern Romania. His results showed that the vegetation diocarbon stratigraphies of lake-sediment and peat se- succession in these mountains broadly corresponds to quences (Farcas et al. 1999; Björkman et al. 2002, 2003; the rest of Romania. Nevertheless, there are some differ- Bodnariuc et al. 2002; Tantau et al. 2003; Tantau 2003; ences in the vegetation succession as compared to other Feurdean, in press), Th-U dating (Onac and Lauritzen areas. 1996; Tamas and Causse 2000) and stable isotope analy- 1. The presence of Fagus pollen grains already in the ses of speleothems (Onac et al. 2002; Tamas 2003). Many Preboreal and rapidly increasing percentages from 2% studies are still ongoing and the obtained results have to 10% during the Boreal let Pop (1942) to assume that not yet been published (Fig. 3). Fagus may have had refugia in the Gutaiului Mountains. 2. Low Carpinus pollen percentages during the Subboreal, as compared to central and southern Roma- Objectives of the doctoral thesis nia were explained as a result of competition with Picea and a strong expansion of Fagus. At the beginning of my PhD project, no radiocarbon- 3. Abies pollen grains are only present in very low num- dated pollen diagrams were available for Romania and bers. very little was known about the vegetation development 4. The expansion of Fagus may have been favoured by during the Lateglacial and Holocene (see e.g., Willis 1994). more moist and cooler climatic conditions during the To address this gap, this thesis focused on a combina- Subboreal, which also led to the formation of peat bogs tion of high-resolution, pollenstratigraphic and plant- in the area. This is well reflected in the stratigraphy by a macrofossil studies of two sites in northwestern Roma- transition from Cyperaceae peat to a peat composed of nia. These analyses were combined with lithological pa- Cyperaceae and Sphagnum (Pop 1942). rameters (loss-on-ignition, mineral magnetics) and a de- tailed AMS 14C chronology.

Figure 3. Topographic map of Romania and location of recently investigated and radiocarbon-dated sites used in the data compilation of Figs. 12-24.

1881 48 SLOVAKIA 22 24 26 28 UKRAINE 48 2061 MOLDOVA

Gu 48 tai ulu i Mt 2305 s. a Study area E astern Tisz Magherus HUNGARY 2102 Iezerul Calimani Carp

Cluj athia thians Apuseni Mts. ns 46 1848 UKRAINE

ROMANIA Mohos tern Carpa s 46 e W Avrig N 2535 Southern Carpathians 2507 Taul Zanogutii 2509 2518 Black Sea

Semenic Mts. Bukarest > 2000 m a.s.l. Danube 1000–2000 44 Danube 400–1000 YUGOSLAVIA 200–400 < 200 BULGARIA 0 50 km 44 22 24 26 28 30

11 Pollen and spore analysis is an important tool for and together they provide more reliable information on vegetation reconstructions. The vegetation is generally past vegetation dynamics and climatic changes. in equilibrium with climate. When climate changes the vegetation tends to respond according to its physiologi- The specific objectives of the doctoral thesis are: cal limits (Iversen 1954; Wright 1984; Ammann et al. 2000; 1.To establish a high-resolution, radiocarbon-dated Wick 2000; Tinner and Lotter 2001; Williams et al. 2002). Lateglacial and Holocene pollen stratigraphy for Therefore pollen and spore analyses can be used to yield north-western Romania (Appendix I, II, III). specific climatic information. Caution is however advised 2.To reconstruct the Lateglacial and Holocene veg- when interpreting changes in the composition of the fossil etation development in north-western Romania, to pollen flora in terms of past vegetation and climatic discuss forest dynamics (arrival, expansion, reduc- changes. Firstly, the main restriction of pollen analysis tion) and to identify which trees might have survived is the identification of pollen at lower taxonomical levels in Romanian refugia during the Last Glacial Maxi- (family and genera). Pollen can rarely be identified at mum (Appendix I, II, III, IV). species level, which makes it difficult to obtain accurate 3.To study whether changes in vegetation develop- palaeocological information (Birks 1973; Watts 1978; ment can be connected to Lateglacial and Holocene Hannon 1999). Secondly, pollen production and the dis- climatic fluctuations and to compare the timing of persion of some pollen types make the interpretation of these to records from around the North Atlantic re- the pollen assemblages in terms of local versus regional gion (Appendix III, IV; Thesis summary). signals difficult. Thirdly, some of the taxa have low pol- 4.To summarize earlier pollen stratigraphic work from len production and/or their pollen grains are very sensi- Romania and to present a plausible scenario for tree tive to corrosion and are thus underrepresented or rarely arrival, expansion and reduction based on recently found in the fossil assemblages (e.g. Larix and Populus). investigated and radiocarbon-dated pollen stratigra- More recently, it has therefore been suggested that pol- phies (Thesis summary). len analysis should always be associated with plant mac- 5.To discuss the impact of humans upon the land- rofossil analysis for a better interpretation of the fossil scape in north-western Romania (Appendix II, III). assemblages, hence, providing more precise information on past vegetation and climatic changes ( Kullman 1998, 2002; Hannon 1999; Birks and Birks 2000; Birks 2003). Study area However, also plant macrofossil analysis has some draw- backs. The lower dispersion capacity of plant macrofos- Climate sils results in fossil remains being more closely depos- ited to the parental plants. Therefore, macrofossils of The modern climate in Romania is continental temperate, terrestrial taxa, particularly of those plants growing more but varies across the country. The north-western part upland e.g., Quercus, Tilia, Ulmus and Corylus are rarely has a rather mild and moist climate, which is influenced found in fossil assemblages, as compared to those grow- by western oceanic air masses, while the eastern part is ing in the close proximity of the sedimentary basin e.g., influenced by cold and dry air masses from the Russian Salix, Betula and Alnus (Birks 1973, 2003; Wainman and plain. The southwest receives warm air masses from sub- Mathewes 1990). Macrofossils of deciduous trees are Mediterranean areas and the southeast is influenced by also very susceptible to decay and are consequently dry air masses from south-western Asia. The Gutaiului strongly underrepresented or rarely found in fossil as- Mountains have a higher amount of precipitation as com- semblages. The lower production rate of macrofossils pared to other regions. The precipitation shows a sig- and their low representation in the sediment in general, nificant altitudinal gradient and ranges from ca. 700 mm/ make it often difficult to perform plant macrofossil analy- yr to ca. 1200-1400 mm/yr at higher elevation. Mean an- sis with a high resolution (Tobolski and Ammann 2000). nual temperature is around 8ºC and mean winter and sum- Pollen and macrofossil analysis complement each other mer temperatures are -3ºC and 12-13ºC, respectively.

Table 2. Percentage representation of the plant species in the Gutaiului Mountains according to their moisture and temperature requirements (Marian, 1999). Moisture Temperature Xerophilous (4,2%) Cryophilous (1,2%) Xeromesophilous (28,22%) Microthermophilous (14,36%) Mesophilous (40%) Micromezothermophilous (61,08%) Mesohygrophilous (15,90%) Moderate thermophilous (10,26%) Hygrophilous (2,56%) Thermophilous (0,51%)

12 23O30‘ Geological map of Gutaiului Mountains (after Edelstain et al. 1980) O 48 Legend Location of study sites Basalts Pyroxene andesite (Gutai type) and biotit dacite (Plesca type) Pyroxene andesite (Sapanta, Mara and Ignis type) and piroxene basaltic andesite (Breze type) Basaltic pyroxene andesites, pyroxene andesites Pannonian and amphibole-pyroxene andesites (Jereapen type) Quartz andesites and quartz-basaltic andesites (Piscuiatu type) Dacites and hyalodacites

Pyroxene andesite (Ilba type)

Sarmatian Basaltic pyroxene andesite and pyroxene andesites

Badenian Rhyodacite proclastic deposit

47O50‘ Sedimentary deposits Quaternary Neogene Palaeogene Intrusions

Veins

Faults 0 5 10 Nappes O ‘ 24O 23 30 Km

Figure 4. Geological map of the Gutaiului Mountains (simplified after Edelstain et al. 1980).

Geology (Istvan et al. 1990). Numerous peat bogs are known within this mountain massif. According to Pop (1960) this area The Gutaiului Mountains belong to the NW Late Pliocene provides ideal conditions (i.e., impermeable bedrock, pre- volcanic arc of the Romanian Carpathians (Borcos et al. cipitation and springs) for the formation of peat bogs. 1979). The bedrock is almost entirely composed of acidic rocks such as andesites rich in pyroxene and quartz, Vegetation dacites, and rhyolites. Sedimentary rocks occur only occasionally (Borcos et al. 1979) (Fig. 4). Brown earth The modern vegetation in Romania is composed of a soils dominate in the mixed-deciduous woodland. Acid mixture of Eurasiatic, central European, circumpolar, At- brown soils are dominant in beech and spruce forests, lantic, Mediterranean and sub-Mediterranean species while podzolic soils are formed under coniferous forests (Donita 1962; Csürös 1976; Cristea 1993). The vegeta-

24 26 Legend N Alpin belt 48 Subalpine belt Figure 5. Map showing the Spruce subbelt Montane vegetation zones (steppe, Beech and Beech- belt spruce subbelt forest-steppe and nemoral Foothill belt zone) in Romania (after Mesophilous oaks Nemoral Thermophilous zone Cristea 1993). The nemoral oaks Forest-steppe zone includes both thermo- Steppe philous and mesophilous Delta 46 oak forests. The figure also shows the distribution of the altitudinal montane for- est belts (foothill, beach- spruce, spruce, sub-alpine

Black and alpine belts). See Fig- Sea ure 6 for comparison.

44

13 m a.s.l. Figure 6. The altitudinal vegetation zones of Romania (after Cristea 1993). The composition of each forest belt is described 2400

in detail in the text. alpine 2200 belt

2000 subalpine belt 1800

1600 t l

e 1400 spruce b belt ne

beech-spruce a 1200 beech-fir subbelt ont m 1000 beech subbelt 800

600 foothill belt 400

nemoral zone forest-steppe 200 steppe zone zone

tion is arranged in zones (latitudinal arrangement) and (2%) (Marian, 1999). The distribution of species accord- forest belts (altitudinal arrangement) according to cli- ing to their moisture and temperature requirements is mate, topographic, and edaphic conditions (Figs. 5, 6). shown in Table 2. Due to western oceanic influences The zones are: steppe zone, forest-steppe zone and and a northern location of the Gutaiului Mountains, the nemoral zone; the latter is the most extensive vegetation altitudinal distribution of the forest belts is as follows: type (Donita, 1962; Csürös 1976; Cristea 1993). The lim- 1. The lowest forest belt (below about 600 m) is domi- its of the altitudinal forest belts vary in the Carpathians nated by sessile oak (Quercus petraea). Other de- with latitude, distribution of the air masses and orienta- ciduous tree species, including common oak tion of the mountain massif. Four altitudinal belts can be (Quercus robur), lime (Tilia cordata), hazel (Corylus distinguished: avellana), ash (Fraxinus excelsior), hornbeam 1. The foothill forest belt (300-600 m a.s.l.) with several (Carpinus betulus) and beech (Fagus sylvatica) oak species (the most common is Quercus petraea), occur occasionally. This belt is affected by human lime (Tilia cordata), hazel (Corylus avellana), horn- activities (e.g. deforestation, forest grazing, and beam (Carpinus betulus) and beech (Fagus cultivation). sylvatica). 2. The beech forest belt (Fagus sylvatica) occurs be- 2. The montane belt is subdivided into three sub-belts: tween 600 and 1000 m. Communities of hornbeam beech (Fagus) between 600-1000 m; beech–spruce (Carpinus betulus), lime (Tilia cordata), hazel (Fagus-Picea) or beech –fir (Fagus-Abies) between (Corylus avellana), ash (Fraxinus excelsior), elm 1000-1200 m and spruce (Picea) between 1200-1800 (Ulmus minor, U. glabra), birch (Betula pendula, B. m. verrucosa), elder (Sambucus nigra) and oak 3. The sub-alpine belt (1800-2000 m) is dominated by (Quercus petraea, Q. robur) are present within the communities of montane pine (Pinus mugo), juniper lower part and spruce (Picea abies) and pine (Pinus (Juniperus communis ssp. nana) and rhododendron sylvestris) occur in the upper part. (Rhododendron kotschyi). 3. Above 1000 m spruce (Picea) is only present as en- 4. The alpine belt occurs above 2000 m and consists claves within the beech forest. Communities of of communities of several willows (Salix spp.) and rowan (Sorbus aucuparia), maple (Acer herbaceous species such as Silene acaulis, pseudoplatanus), grey alder (Alnus incana) and fir Saxifraga bryoidis, Festuca glacialis, Sesleria (Abies alba) occur sporadically within this belt. coerulans and Carex curvula. 4. The sub-alpine belt is only present on the highest The modern flora in the Gutaiului Mountains contains a peaks of the mountain massif surrounding the study mixture of Eurasiatic (43,9%), European (13,5%) and cen- area and consists of communities of dwarf pine tral European (9,69%) taxa as an expression of a conti- (Pinus mugo), juniper (Juniperus communis ssp. nental temperate climate. Additionally, there are circum- nana), rhododendron (Rhododendron kotschyi), polar (11,2%), Atlantic (2,4%), Mediterranean and sub- lingonberry and blueberry (Vaccinium vitis idea, V. Mediterranean (1,5%) and Mediterranean-pontic species myrtillus) and green alder (Alnus viridis).

14 23o 23o30' N y U k r r a i 48o a n e W E g 824 T is n a S u H Figure 7. Map over north- Tur 587

G western Romania and lo- u t a cation of the surveyed Negresti-Oas iu l u sites in the Gutaiului i 1240 M o Iz Mountains. The position Satu Mare u a n S ta of the two investigated om 1200 i es ns sites Steregoiu and Preluca Izvoare Tiganului is shown by the 1307 black square, and the lo- Creasta Cocosului 987 cation of the sites Izvoare L 1443 a Baia Mare p u and Creasta Cocosului is s marked with black dots.

0 50km 47o30'

Initially, several peat bogs in the Gutaiului Moun- Material and methods tains were surveyed and cored. However, most of them turned out to be young peat deposits and did not meet Fieldwork and site selection strategy our requirements. Parallel cores with 50 cm overlap were retrieved from the basal sediments of two former lake The study area was chosen in accordance with the aims basins: Creasta Cocosului (N 47°40’; 23°50’; 900 m a.s.l.) outlined above. There is no published data about the and Izvoare (N 47°44’80’’; E 23°43’35’’; 900 m a.s.l (Fig. presence of glaciers during the Weichselian, but it is 7). The lithostratigraphic description of the sediments is generally assumed that alpine glaciers did not reach be- presented in Tables 3 and 4, respectively. A preliminary low 1600-1800 m a.s.l. in the eastern Carpathians analysis showed a dominance of thermophilous tree (Woldstedt 1958; Balteanu et al. 1998). Therefore it was pollen (Quercus, Tilia, Ulmus), characteristic of an early expected to obtain continuous and old sediments (Late to mid-Holocene deciduous forest. Therefore, these sedi- Glacial Maximum to present) in the basins located below ments were regarded as too young for the purpose of this elevation. A more recent study by Kortarba and the present investigation. Baumgart-Kortarba (1999) gave evidence for the exten- At the end, two sites were selected for this PhD sion of glaciers during the Last Glacial Maximum at project: the former small crater lakes Preluca Tiganului around 950 m in the Polish Carpathians, which could (N 47°48’83’’; E 23°31’91’’; 730 m a.s.l.) and Steregoiu (N imply that the study area in the Gutaiului Mountains 47°48’48’’; E 23°32’41’’; 790 m a.s.l.), situated on the may have been subjected to periglacial processes. western flank of the Gutaiului Mountains (Appendix I,

Table 3. Lithostratigraphic description of the Creasta Cocosului sediment cores between 7.88-6.35 m below surface (The boundaries between different layers were gradual). Units Depth (m) Description 3 6.35–7.72 Gyttja, dark brown 2 7.72–7.82 Gyttja clay with sand, dark brown 1 7.82–7.88 Silty clayey gyttja with sand, rich in diatoms, dark brown

Table 4. Lithostratigraphic description of Izvoare sediment cores between 3.82-2.25 m below surface (gLB, gradual lower boundary; sLB, sharp lower boundary). Units Depth (m) Description 4 2.25–3.01 Sphagnum peat, gLB 3 2.82–2.88 Peat, gLB 2 2.88–2.96 Gyttja, dark bown, gLB 1 2.96–3.82 Sandy silt with sand layers and gravel, grey, sLB

15 Foto 8

Figure 8.View over the former crater lake Preluca Tiganului Figure 9. View over the site Steregoiu

II, III and IV). Cores were collected with a Russian corer much detailed palaeoecological information as possible. (1 m length, 5 cm diameter) from the central, deepest part of each site in May 1999 (Figs. 8, 9). Overlapping cores Laboratory methods were taken in order to obtain enough material for bios- In the laboratory the cores were described in detail (Ap- tratigraphic and lithostratigraphic analyses. The cores pendix I, II, III) and sub-sampled for pollen, macrofossil, were preliminary described in the field, wrapped in a plas- loss-on-ignition (LOI) and mineral magnetic analyses. tic film and transported to the Department of Geology at Lund University, Sweden where sub-sampling and labo- Pollen and charcoal analyses ratory work were performed. Pollen analysis was initially carried out at 2-cm intervals The reason for studying two neighboring sites (ca. 1 at Steregoiu. Later, micro-charcoal particles were counted km from each other) was to see whether the vegetation on each second pollen slide, i.e. every 4th cm. At Preluca development reconstructed at one site can be corrobo- Tiganului the resolution between analysed samples var- rated by the second site. Both sites are former crater ied between 2.5 cm in the Lateglacial and early Holocene lakes and their stratigraphies may be subject to hiatii, sediments and 4 cm intervals for the mid- and late rendering a continuous reconstruction impossible. We Holocene. Pollen samples (1cm3) were prepared accord- speculated that the location of the sites at 700-800 m ing to the standard procedure described by Berglund a.s.l. and in deep-incised valleys may have offered suit- and Ralska-Jaseiwiczowa (1986) and Moore et al. (1991). able climatic conditions for tree survival during the Last Lycopodium tablets with a known number of spores were Glacial Maximum (see discussion on glacial refugia). added to each sample to determine fossil pollen and char- Moreover, middle to high altitude sites are documented coal concentrations (Stockmarr 1971). Pollen counts were to have been more sensitive to vegetation changes dur- generally made at 400x magnification but for special iden- ing the cold phases of the Lateglacial, due to their pos- tification a 1000x magnification and oil immersion was sible location at an ecotone i.e. boundary zone between used. Pollen and spore identification was based on keys two different vegetation belts (Peteet 2000; Wick 2000; and illustrations of Moore et al. (1991) and Reille (1992, Ammann et al. 2000). Preluca Tiganului (surface area ca. 1995), and by comparison to reference collections at the 1 ha) and Steregoiu (surface area ca. 0.5 ha) are topo- Department of Geology, Lund University. An average of graphically closed basins where the hydrological bal- 500 pollen grains, excluding aquatics and spores, were ance is controlled by temperature (T), precipitation (P) counted at each level, and 95 pollen taxa were indentified and evaporation (E) changes and may therefore be used in total. Percentages of terrestrial pollen were calculated for a reconstruction of lake level fluctuations and past on the basis of their total sum, excluding spores and climatic conditions. Since the source area for pollen is pollen of aquatics. Percentages of spores and aquatic related to the size of the basin, these small basins should pollen types were calculated on the basis of the total give a local environmental signal (Jacobson and sum including terrestrial, spores and aquatic pollen types. Bradshaw 1981; Sugita 1994; Broström 2002). We chose The pollen diagrams were drawn using TILIA and TILIA to perform a detailed multi-proxy analysis of two sites Graph (Grimm 1992). To facilitate the description and in- (pollen, macrofossil, mineral magnetic and organic mat- terpretation of the pollen diagrams, the pollen spectra ter analyses, radiocarbon dating), instead of analyzing were divided into local pollen assemblage zones (LPAZ) several sites by one proxy method only to obtain as using the CONISS program (Grimm 1987).

16 Macrofossil analysis Results Contiguous (4 cm thick) samples were taken for macrofossil analyses from both sites. The samples from Summary of papers I – IV Preluca Tiganului have, however, only been analysed in the lowermost 6 m, comprising the time interval between Paper I 14,500 and 8000 cal. yr BP. The sub-samples were soaked Björkman L, Feurdean A, Cinthio K, Wohlfarth B, in 5% NaOH and sieved through a 0.25 mm mesh under Possnert G. 2002. Lateglacial and early Holocene veg- running water. All recognizable macro remains were ex- etation development in the Gutaiului Mountains, NW amined under a dissecting microscope. The fossil identi- Romania. Quaternary Science Reviews 21: 1039-1059. fication was carried out using atlases and keys of Katz et al. (1965). The results are presented as volume percent- This is the first paper presenting the vegetation de- ages. The macrofossil diagrams were also constructed velopment in the Gutaiului Mountains and the response with the TILIA program (Grimm 1992) and visually sub- of the vegetation to Lateglacial and early Holocene cli- divided into macrofossil assemblage zones (MAZ). matic changes. The reconstruction is based on pollen- Betula sect. Albae is referring to remains of tree birch stratigraphic investigations of the basal sediments from which could not be identified to species level. the two small, overgrown crater lakes Preluca Tiganului (730 m a.s.l.) and Steregoiu (790 m a.s.l.). This study builds Mineral magnetic analysis on work by Wohlfarth et al. (2001), which only covered The sequences were continuously sub-sampled (2 cm3) the early part of the last deglaciation (15,000- 13,600 cal. for mineral magnetic susceptibility and Saturation Iso- yr BP), but showed the potential of these sites for fur- thermal Remanent Magnetization (SIRM). The analyses ther investigations. The main aim was thus to provide a were used to assess input of minerogenic material and continuous record of the spatial and temporal vegeta- soil erosion. The samples were dried over night at 40 ºC tion development, based on high-resolution pollen stud- prior to mineral magnetic analyses. Magnetic suscepti- ies from an area where little knowledge had previously bility was measured in a low magnetic field of 0.1 mT been available. Another aim was to establish a detailed using a balanced alternating current bridge circuit. Mass chronology for Lateglacial vegetation changes in a re- specific units were calculated and expressed as mm3kg-1. gion, where virtually no radiocarbon dates had been SIRM was induced in a strong magnetic field of 1 Tesla obtained. This new study also aimed at advancing by a Redcliff BSM 700 Puls Magnetic Charger. The re- palaeoecological knowledge from a key area, which has sulting remanent magnetization was measured with a been regarded as a potential glacial refugia during the Molspin Spinner Magnetometer. Mass specific units last cold stage of the Quaternary. were calculated as mAm2kg –1. The pollen record shows that the open vegetation with herbs (Artemisia, Chenopodiaceae and Poaceae) Loss on ignition (LOI) and shrubs (Salix and Juniperus) dominated the sur- Loss on ignition (LOI) was measured on the same sam- rounding areas prior to 14,700 cal. yr BP, corresponding ples on which we performed mineral magnetic analyses, to the end of the Late Glacial Maximum. Marked warmer following the methods described by Bengtsson and Enell conditions at the beginning of the Lateglacial interstadial (1986). The samples were only combusted at 550ºC for 3 led to a rapid expansion of an open forest dominated by hours, because previous research has shown that car- Betula and Pinus. The renewed spread of open vegeta- bonates are not present in the sediments (Wohlfarth et tion with Artemisia, Poaceae, Cyperaceae, and al. 2001). LOI is expressed as percentage of the weight Chenopodiaceae between 14,050 and 13,800 cal. yr BP of the dried sample and was used for estimating the or- suggests colder climatic conditions which could corre- ganic matter content and help to characterize the sedi- spond to GI-1d in the GRIP event stratigraphy. A further ment composition. expansion of a more diverse and dense forest with Betula, Picea, Pinus, Alnus and Ulmus from 13,400 cal. yr BP Radiocarbon dating onwards indicates a rise in temperature which coincides The chronology is based on seventeen AMS radiocar- approximately with GI 1a-1c. This was followed by an- bon dates from Steregoiu and on fourteen AMS 14C meas- other episode of forest reduction between 12,900 and urements from Preluca Tiganului. All measurements were 11,500 cal. yr BP, implying cold climatic conditions. The performed on terrestrial plant macrofossils and/or on peat. expansion of Betula, Alnus and Pinus dominated, open The obtained AMS radiocarbon ages were converted forests at 11,500 cal. yr BP, and of Picea and Ulmus at into calibrated years BP using the calibration curve of 11,250 ca. yr BP suggests a climate warming at the begin- Stuiver et al. (1998) and the OxCal v3.5 program (Bronk ning of the Holocene. Ramsey 1995). This study allowed for the first time a comparison between environmental changes recorded during the Lateglacial and early Holocene in Romania with those

17 reconstructed from well-dated sequences in central and The local establishment and expansion of Quercus, Tilia, northwestern Europe. The reconstructed vegetation suc- Fraxinus and Acer was dated to ca. 10,750 cal yr BP. The cession revealed that the climatic fluctuations recorded late expansion and establishment of Quercus, Tilia, in the North Atlantic region had a distinct impact on Fraxinus and Acer may imply that these trees survived areas situated in the continental part of Europe e.g., Ro- in glacial refugia in areas located further away e.g. south mania. The major conclusion of this paper is that climatic of Romania and in the Hungarian plain. At ca. 10,200 cal. fluctuations may have been the main factor controlling yr BP Corylus started its expansion and replaced Ulmus the forest composition during the Lateglacial in Roma- as a dominant forest constituent. A forest composed of nia, while the early Holocene forest composition was Corylus, Ulmus, Ouercus, Tilia, Fraxinus, Acer and Picea primarily determined by succession and migration proc- was maintained in the surrounding landscape until ca. esses. 4800 cal. yr BP. It was succeeded by a simultaneous es- tablishment and expansion of Fagus and Carpinus at Paper II around 4800 cal. yr BP. From about 4000 cal. yr BP on- Björkman L, Feurdean A, Wohlfarth B. 2003. wards Fagus dominated and the forest diversity became Lateglacial and Holocene forest dynamics at Steregoiu strongly reduced, with only Quercus and Carpinus in the Gutaiului Mountains, NW Romania. Review of present. The late Fagus and Carpinus expansion was Palaeobotany and Palynology 124, 79-111. probably connected to climatic changes, rather than human activities. Human activities are represented by This article presents a detailed vegetation history, based grazing and forest grazing. Scattered pollen indicators on the sequence from Steregoiu, covering the time pe- of arable field and grazing are recorded during the last riod from the end of Last Glacial Maximum to the present 300 years. These pollen grains were probably blown from day. The objectives were to obtain a high-resolution time the Oas Depression, which is located at an elevation of scale for vegetation changes and a chronology for tree 200 m above sea level. dynamics. The discussion is focused on the possible survival of tree taxa in the study area, on immigration Paper III routes, tree dynamics (arrival, expansion, maximum and Feurdean A. (in press). Holocene forest dynamics in retraction) and human activities. The article also presents north-western Romania. The Holocene. the on-site vegetation development and the sedimen- tary history of the basin, helping in the interpretation of The aim of this study was to document the postglacial the depositional environment. It is the first most com- forest development based on a pollen-stratigraphic analy- plete pollen and radiocarbon-dated sequence presented sis of the sediment sequence obtained from Preluca so far from Romania. This facilitates comparison to other Tiganului, north-west Romania. The terrestrial pollen data pollenstratigraphic studies from Romania and surround- was used to describe the history of the upland vegeta- ing regions, allowing for a regional reconstruction of the tion. Aquatic and wetland pollen types were used to vegetation history and climatic. reconstruct past peat surface conditions, which in turn The results show that the cold periods that occurred may provide an estimate of climatic conditions. Further- prior to 14,700 cal. yr BP and between 12,900 and 11,500 more, results obtained at Preluca Tiganului were corre- cal. yr BP were characterized by the presence of open lated to those obtained at Steregoiu in order to infer vegetation communities with low shrubs (Salix and regional forest and climate history. Apart from the timing Juniperus) and herbs (Artemisia, Poaceae, of the postglacial tree succession, ecological require- Chenopodiaeae). The forest part of the landscape was ments e. g. tolerance limits, length of the growing sea- vegetated by Pinus and Betula, although high amounts son, regeneration, shade tolerance, colonization ability, of Pinus pollen were possibly derived from open for- modern distribution of the trees and inter-specific com- ests, which were restricted to lowlands areas. During petition of different taxa, as well as factors that may have warmer episodes between 14,700-14,050 cal. yr BP and influenced the Holocene tree succession, are discussed. 13,750-12,900 ca. yr BP, the open vegetation was replaced The first step in the Holocene vegetation succession by open forest composed of Betula, Picea, Pinus and was the rapid expansion of birch and pine, which sur- Alnus. Small Ulmus populations may have existed lo- vived in the area during the Lateglacial. During the first cally between 13,200 and 12,900 cal. yr BP. 200-300 years of the Holocene, the forest had an open The rapid expansion of open Betula, Pinus and Alnus character and large areas of the landscape were still oc- forest at the beginning of the Holocene, at 11,500 cal. yr cupied by an open shrub and herb vegetation. The BP is the result of the expansion from local populations. slightly delayed expansion of Ulmus and Picea (at ca. The fairly abrupt replacement (at ca. 11,250 cal. yr BP) of 11,250 cal. yr BP) was connected to a strong reduction pine-birch forest by Ulmus and Picea indicates a tem- during the cold Younger Dryas episode. The arrival of perature rise and that they had occurred in small stands Quercus, Tilia, and Fraxinus probably occurred around in isolated habitats with sufficient warmth and humidity. 11,250 cal. yr BP, however, they did not expand until

18 around 10,700 cal. yr BP. Their delayed expansion was This paper combines results from high-resolution primarily connected to the immigration lag from more macrofossil, pollen, mineral magnetic and loss-on-igni- distant localities as compared to the other tree taxa. The tion analyses from two sites in north-western Romania. demand of higher temperatures during the growing sea- The aim of this study is to complement previous pollen son and high growing degree days have played an im- results by focusing on the reconstruction of the local portant role in their spreading from 10,700 cal. yr BP. In vegetation and on climatic conditions. Terrestrial plant contrast, competition with the already established Ulmus, macrofossils were used for a reconstruction of the local Picea, Pinus and Betula may have limited their expan- vegetation and telmatic and limnic macrofossils, includ- sion. ing faunal and flora remains, were used to document Temperature reconstructions based on a variety of past lake level fluctuations and to improve the knowl- proxies show general warming trends during the early edge about climatic changes. Lithostratigraphy and min- Holocene, which were interrupted by short, colder epi- eral magnetic parameters (magnetic susceptibility and sodes. The changes in temperature and precipitation were SIRM) can give information about water-level fluctua- significant enough to trigger changes in forest composi- tions and erosion of sediments from the surrounding tion. This study presents for the first time a correlation slopes. The organic matter content and the composition between changes in forest composition in Romania and of the aquatic remains were used as an approximate indi- short-lasting climatic cooling episodes recorded in the cator of aquatic productivity and of changes in the Northern Hemisphere. For instance the pollen investiga- depositional environment. tions show that the permanent decline of Ulmus along The cold water lake with low trophic status was colo- with the expansion of Corylus at ca. 10,300 cal. yr BP, the nized by pioneer communities prior to 14,700 cal. yr BP. reduction of Ulmus, Tilia and Quercus together with the The prevalence of a cold and dry climate is indicated by dominance of Corylus at ca. 9300 cal. yr BP, and a clear the presence of a tree-less vegetation and by the devel- decline of all thermophilous trees taxa at 8600 cal. yr BP opment of open communities including shrubs and herbs may possibly be synchronous with the short climatic with steppe and montane taxa. Relative stable environ- cooling recorded around 10,300 cal. yr BP, 9300 cal. yr BP mental conditions characterized the beginning of the and 8600 cal. yr BP in regions around the North Atlantic. Lateglacial as revealed by the formation of open forests Once Ulmus started to decline (around 10,300 cal. yr BP), with Pinus and Betula, and changes in sediment compo- it never regained its dominance in the forest. Instead, sition (reduction of the mineral magnetic material and Quercus, Tilia and Fraxinus re-expanded around 8200 rise in organic matter content). The increase in number cal. yr BP, although except for Quercus, the latter two of limnic animals could reflect some rise in lake produc- taxa never reached their previous proportion in the for- tivity. This succession was interrupted by a short epi- est again. sode of climate cooling and drier conditions inferred from At Preluca Tiganului we have for the first time dated macrofossils, which led to a forest reduction with single the local establishment of Carpinus and Fagus. Thus, stands of Pinus and Betula between 14,100 and 13,800 the expansion of Carpinus at ca. 5750 cal. yr BP, coinci- cal. yr BP. Plant macrofossils provide evidence of a rapid dent with a regression of Picea and Corylus was docu- expansion and dominance of Picea abies forests mixed mented to be induced by warmer and possibly drier cli- with Pinus sp., Larix decidua, Betula sp., and Salix sp. mate. The expansion and dominance of Fagus at 5200 from about 13,800 cal. yr BP, while pollen data indicates cal. yr BP and 4800 cal. yr BP, respectively is interpreted a possible occurrence of Ulmus. Warm and moist cli- as a response to cooler summers, warm winters and moist matic conditions were inferred between 13,800 and 13,400 conditions. An alternative cause for the late expansion cal. yr BP, followed by a decrease in humidity from ca. of Carpinus and Fagus is human impact, although evi- 13,400 cal. yr BP. Cooler and drier conditions between dence of human-induced deforestation is not clearly 12,900 and 11,500 cal. yr BP are inferred from the recur- documented. Only during the last 300 years is human rence of an open landscape where only Betula, Larix, activity clearly recognized by a strong reduction of for- Salix, and possibly Pinus were present. est diversity and density. Other evidence of human ac- An increase in diversity and abundance of limnic plant tivities upon the landscape are grazing, forest grazing and animal remains around 11,800 cal. yr BP (most pro- and agriculture, the later only evidenced in the Oas De- nounced at 11,500 cal. yr BP) reflects higher water tem- pression. peratures and increased aquatic productivity. At 11,500 cal. yr BP, terrestrial macrofossils show a rapid reforesta- Paper IV tion with Pinus, Betula, Larix, Populus and Picea. This Feurdean A. and Bennike O. Late Quaternary is taken to mark the rapid warming at the beginning of palaeocological and paleoclimatological reconstruction the Holocene, but is 300 years later than the signal ob- in the Gutaiului Mountains, NW Romania. Manuscript served in the aquatic community. Reconstructed climatic submitted to Journal of Quaternary Science. conditions for the early Holocene were warm and wet, but were followed by a drier climate from about 10,600

19 ) cal. yr BP onwards. This observation is also supported P B s 5 r 2 by the dominance of trees with more continental affini- s ea ) > te y s s s a . (m al u u lu d l o c n s a y ca th ties such as Quercus, Tilia, Ulmus and Fraxinus be- er xi u e r C ( p arc u lia a lm ic o 4 e e h i r P C 1 g D C tween 10,700-8600 cal. yr BP, which indicate higher sum- Q T F U A Particles/cm3 x 1000 0 0,0 50 +75 mer temperatures and lower precipitation as compared 100 200 300 400 500 500

1000 to the present day. Dry summers (inferred from aquatic 1180 +75 1500 0,5

macrofossils) could have been responsible for the late 2100 +75 2000 expansion and maximum of Corylus at 10,300 cal. yr BP 2500 3000 1,0 3130 +75 and 9300 cal. yr BP, respectively. Very dry conditions 3500

3680 +70 4000 may have occurred between 8600 and 8200 cal. yr BP. 4500 5000 5500 1,5 The clear reduction of Tilia, Quercus, and Fraxinus 6000 6500 7000 around 8600 cal. yr BP may have been caused by in- 6425 +75 7500 creased water stress and possibly cooler conditions. 2,0

Moisture availability increased again after 8200 cal. yr 8000 2,5 BP, documented by the expansion of Picea and by the 7560 +80

7655 +85 re-expansion of Quercus, Tilia, and Fraxinus and by 8500 wetter peat surface conditions. 7870 +75 3,0 9000

8300 +85

8215 +85 9500 3,5 Unpublished results 10000

9130 +95 4,0 Additional results to those presented in Appendix I-IV 10500 9530 +85

11000 are variations in micro-charcoal content along the whole 9665 +110 4,5 sediment sequence at Steregoiu (Fig. 10) and a mid- to 10325 +150 11500 12000 late Holocene macrofossil record for Steregoiu (Fig. 11). 12500 5,0 13000 10910 +105 Micro-charcoal was analysed to obtain additional pa- 13500 12365 +115 14000 14500 rameters on forest fires and past climatic conditions and 5,5 the late Holocene macrofossil analysis contributes to a

reconstruction of the terrestrial local vegetation and al- 6,0 20 20 20 40 20 40 20 40 lows extending the reconstructed peat surface status 100 200 300 400 500 from 8000 cal. yr BP to the present. Figure 10. Percentage diagram showing selected tree pollen Fire history taxa and micro-charcoal concentrations in the sequence from Prior to 14,700 cal. yr BP, micro-charcoal concentrations Steregoiu. The diagrams are presented on a linear depth scale are generally low, indicating that fires have not been and calibrated years are shown on a non-linear timescale. Ra- common around the site (Fig. 10). The slight increase of diocarbon dates are marked by black bars and are given to the micro-charcoal fragments between 14,700 and 13,400 cal. left of the calibrated timescale. yr BP could possibly indicate dry climatic conditions, which could have facilitated vegetation inflammability. The distinct rise in charcoal concentrations from about bers. This could indicate that forest fires decreased and 13,400 cal. yr BP onwards is associated with a decrease that summers were less dry. On the other hand, low char- in pollen values for Picea. The charcoal particles could coal numbers could also indicate that the open vegeta- not be identified, but it could be speculated whether tion communities, which became established between intense forest fires could have destroyed parts of the 12,500 and 11,700 cal yr BP, were less susceptible to igni- Picea stands around the site. A clear correlation between tion. high numbers of macro-charcoal and a reduction of Picea The increase in micro-charcoal particles between pollen and macro-remains could e.g., be demonstrated 11,700 and 11,000 cal. yr BP occurs at a time when dense by Wohlfarth et al. (2001) at Preluca Tiganului for the forest became gradually established around the site. The time interval 13,900-13,700 cal yr BP and was interpreted pollen records show no marked changes in the arboreal as having been caused by dry summer conditions. High pollen percentage values, which may imply that the char- micro-charcoal percentages and increasing pollen val- coal fragments were possibly derived from ground fires ues of xerophytic herbs (e.g Artemisia, Chenopodiaceae, close to the site, which did, however, not affect the for- Poaceae) between 12,900 and 12,500 cal. yr BP could est. The distinct peak in micro-charcoal concentration point to dry summers, which could have caused forest between 10,600 and 10,400 cal. yr BP follows after fires. Herb pollen values remain high between 12,500 and Quercus, Tilia and Fraxinus had become established 11,700 cal. yr BP, but charcoal particles decrease in num- and predates the forest reorganization, (expansion of

20 Steregoiu m ilu h s p a u o ri e e a a a g ct lm p ) e s a u ro m s r u c u . . s cm u e e e e la e p sp e ( cc i a a io u p s t h o b ill id d d s s s ll a t c a t s n u ru d p o a n u a e p x e ice C e n te b ic p o re p d 4 D e ice o u rt ili yc a y a 1 C P P R U F L C C R MAZ 0 50+75 S13

1180+75 50 2100+75 S12

100 3130

3680+70 S11 150

S10 6425+75

200 S9

100 200 300 400 20 40 60 20 40 20 40 20 40 60 80 250

Ole Bennike Figure 11. Macrofossil concentration diagram for Steregoiu with all identified macro-remains presented on a depth scale. 14C dates are shown to the left. The macrofossil zone (MAZ) numbers follow the macrofossil assemblage zones S1-S8 presented in Appendix IV.

Corylus) by ca. 100-200 years. Although no decline in and Urtica dioeca remains (Fig. 11). arboreal pollen percentages and thus in the proportion MAZ-S10 (1.90-1.50 m; 7500-5500 cal. yr BP). of forest taxa can be observed, small, local fires may Macrofossils of Picea abies disappear, although pollen have occurred on or close to the site, but they did not values remain constant. Radicells increase and a few iso- affect the forest composition. Charcoal concentration lated remains of Carex sp. are recorded. Wetter condi- decreases gradually until ca. 10,000 cal. yr BP, indicating tions on the surface of the peat bog could have caused only occasional forest fires. Two isolated charcoal peaks a retraction of Picea, but most probably it was still abun- can be seen at ca. 9300 cal. yr BP and at ca. 8700 cal. yr dant in the surrounded landscape. BP, which may point to small local fires. From 8600 cal yr MAZ-S11 (1.50-1.20 m; 5500-3700 cal. yr BP). Picea BP onwards, charcoal concentrations are generally low, abies remains re-appear and increase gradually; Carex except for between 5700 and 4800 cal. yr BP, 3200 and sp. remains are rare. Cenococcum geophilum sclerotia 2700 cal. yr BP, 1300 and 800 cal yr BP and during the last increase distinctly in the lower part of the unit, whereas 300 years. This could be interpreted as several small lo- radicells are only present in low numbers. This develop- cal fires, which occurred in the surroundings or on the ment, together with the increase in pollen values for site. Picea, Alnus and Betula, could indicate that woody spe- cies expanded over the bog surface, which in turn could Macrofossil results point to a change towards drier peat surface conditions. The macrofossil assemblage zones (MAZ) from MAZ-S12 (1.20-0.30 m; 3700-800 cal. yr BP). The dis- Steregoiu are shortly presented below. The zone num- appearance of Picea abies remains and the presence of bers follow the MAZ S1-S8 presented in Appendix IV abundant Cenococcum geophilum sclerotia coincide and their age assignment is according to Appendix II. with the appearance of radicells and a marked increase MAZ-S9 (2.30-1.90 m; 8000-7500 cal. yr BP). Picea of Carex spp. and Cyperus sp. abies macro remains are abundant, coinciding with high MAZ-S13 (0.30-0.00 m; 800 cal. yr BP to present). pollen percentages for Picea. This indicates that the Potentilla erecta macrofossils increase in numbers, co- surrounding landscape was predominantly forested by inciding with high pollen values for Potentilla. Finds of Picea, but also that this tree likely grew on the peat Cyperus sp. are recorded in the lower part of S13, where surface. Radicells are scarce and Cenoccocum geophilus also radicells decrease markedly. The macrofossil remains sclerotia appear only as two isolated spikes. The telmatic of MAZ-S12 and MAZ-S13 show that fen communities flora is scarce and comprises Carex ssp., Cyperus sp. dominated the surface of the bog. 21 Discussion Forest development in the study area during last ca. 18,000 cal. yr BP and Glacial refugia comparison to other sites in Romania

The term glacial refugia is used to denote areas with Last Glacial Maximum (LGM) favorable micro-environmental conditions, where trees Pollen and plant macrofossil remains in the oldest sedi- could survive in small populations during glacial peri- ments at Steregoiu, which were dated to >14,700 cal yr ods (Faegri 1963; Lang 1970; West 1977; Huntley and BP, i.e. to the end of the LGM, suggest predominantly Birks 1983; Bennett et al. 1991; Willis 1992a, 1992b, 1996, tree-less vegetation with herbs and shrubs of montane 1994). The information about the location of refugia was and steppe character at mid-altitude sites in the Gutailului based on the assumption that the first tree taxa, which Mountains. However, it cannot be excluded that scat- responded to climate warming, were those that had been tered Pinus and Betula individuals were present (Ap- locally present earlier (West 1977; Huntley and Birks pendix IV). Two other pollen-stratigraphic records from 1983; Bennett et al. 1991). Refugial areas offered specific Romania, Iezerul Calimani in the eastern Carpathians local conditions (i.e., moister, warmer) for trees to sur- (Farcas et al. 1999; Farcas 2001) and Avrig in central vive and acted as a repopulation centre for temperate Romania (Tantau 2003), show LGM sediments in the basal trees, which had become extinct in regions, where full part of the sequence (Figs. 3, 12). One AMS 14C date of glacial and/or periglacial conditions prevailed. Once a 14,800±1100 yr BP (ca. 18,000 cal. yr BP) from Iezerul tree species also became extinct in its refugia at any time Calimani may give indications for a LGM age of the sedi- of the Quaternary period, it became extinct everywhere ments, which contained relatively high frequencies of in Europe (Bennett et al. 1991). Pinus pollen, together with some Betula and Picea It has been proposed that tree refugia may have been (Farcas et al., 1999, Farcas, 2001). However, due to un- located in southern Europe, particularly in the Balkans, certainties in the radiocarbon chronology, it has been Alps, Carpathians and the Italian mountains (van der suggested to attribute this part of the sequence to the Hammen et al. 1971; Huntley and Birks 1983; Bennett et Lateglacial interstadial (Farcas et al., 1999). The lower- al. 1991; Willis 1994). The Iberian Peninsula was often most sediments at Avrig are dated to 13,880±90 14C yr BP excluded due to extreme dry condition during glacial pe- (ca. 17,000 cal. yr BP) and contained pollen of Pinus, riods. The Balkan region was situated south of the major Betula, Juniperus and Salix (Tantau, 2003). Recently, a ice sheets and outside the region of continuous perma- pine log was found at Magherus, a site situated in the frost during the Weichselian glaciations and local gla- lowland area, ca. 200 km south of the Gutaiului Moun- ciers may only have occurred at high elevation sites in tains (Fig. 3) and was dated to ca. 14,500 14C yr BP (cor- the Carpathians. Nevertheless this region had one major responding to ca. 17,000 cal yr BP) (Wohlfarth et al. un- drawback: moisture availability. Therefore, if refugia ex- published; Lascu, 2003), demonstrating the existence of isted on the Balkans, they would have been restricted to pine in this area. mid-elevation sites, where adequate habitats may have No published palaeotemperature reconstructions for been available and where potential glacial refugia may Romania exist, but simulated climatic conditions for Eu- have existed (Willis 1994). rope during the LGM show that mean January tempera- High pollen values for coniferous (Picea, Pinus, tures may have been around -15 to -20°C and mean July Abies, Larix) and temperate trees (Betula, Alnus and temperatures at around 13-15°C (Renssen and Isarin, Salix) are often associated with macro-remains during 2001). Permafrost conditions existed in areas north of 47° the Last Glacial Maximum and the Lateglacial confirming N (Renssen and Isarin, 2001). Another paleoclimatic simu- their local presence. In contrast, sporadic finds or low lation for central Europe shows similar low temperatures amounts of pollen of thermophilous trees (Ulmus, and implies dry climate conditions (Kutzbach et al. 1993). Quercus, Tilia, Fraxinus and Corylus) are difficult to Pinus and Betula trees could survive these low tem- interpret in terms of local presence or absence. If it is peratures, but must have been exposed to water stress, assumed that these trees only survived in small popula- caused by drought (less precipitation) or by frozen soils. tions or as single trees, it cannot be expected to find a The pollen and plant macrofossil studies described high amount of pollen grains and macro remains in the above suggest a possible occurrence of Pinus, Juniperus, sediments. Moreover, pollen trap studies show that the Betula and Salix at the end of the LGM in different re- pollen production, especially Pinus is controlled by gions of Romania, such as the Gutaiului Mountains, changes in climate (Hicks 2001; Autio and Hicks 2004). Calimani Mountains and Avrig (Fig. 12). Based on find- Colder climatic conditions could therefore have contrib- ings of pine megafossils it can, however, be concluded uted to a weaker representation of thermophilous pollen that pine forests have been common in lowland areas in in the fossil assemblages. Romania (Wohlfarth et al. unpublished). Picea pollen (ca. 4%) has only been found at Iezerul Calimani and is absent in the lowermost sediments at Steregoiu

22 SLOVAKIA 22 24 26 28 UKRAINE Key MOL

DOV Pinus 48 A

HUNGARY Betula Gutaiului Mts. Salix Magherus Iezerul Calimani Juniperus Cluj 46

46

YUG Avrig

OSL

A VI A Taul Zanogutii Bukarest Black Sea

44

18,000-14,700 cal. yr BP BULGARIA 0 50 km 44 22 24 26 28 30

Figure 12. Map showing the tree taxa, which were present in Romania during the Last Glacial Maximum, according to pollen and mega-fossil evidence (Appendix I, II, IV; Wohlfarth et al. 2001; Wohlfarth unpublished; Lascu 2003) or only pollen (Farcas et al. 1999; Tantau 2003).

SLOVAKIA 22 24 26 28 UKRAINE Key

48 Pinus HUNGARY Betula Gutaiului Mts. Magherus Salix Iezerul Calimani

Cluj Juniperus 46

46 YUG Avrig OSL

A VI A Taul Zanogutii Bukarest Black Sea

44

14,700-14,100 cal. yr BP BULGARIA 0 50 km 44 22 24 26 28 30

Figure 13. Map showing the tree taxa, which were present in Romania between 14,700 and 14,100 cal. yr BP. Their occurrence is based on pollen and macrofossil remains (in the Gutaiului Mountains) (Appendix I, II, IV; Wohlfarth et al. 2001), on charcoal at Magherus (Wohlfarth unpublished; Lascu 2003) and on pollen (Tantau 2003; Farcas et al. 1999).

23 (Gutaiului Mountains) and Avrig. However, the age as- tains differs slightly from other sites in Romania (Fig. signment of the Iezerul Calimani profile is rather uncer- 15), where high pollen values of Pinus and increasing tain and the pre-Holocene pollen assemblages are rela- percentages for Betula and Picea were interpreted as tively homogenous. On the other hand, pollen and char- Pinus-Betula dominated forests, which included some coal analyses could confirm that trees of Pinus sylvestris, Picea and Salix (Farcas et al. 1999; Tantau 2003). Mod- Pinus cembra, Picea sp., Betula, Juniperus sp., Salix ern studies on Picea pollen productivity, dispersion, sp. and Larix, as well as Carpinus betulus existed in transport and representation have, however, shown that Hungary between 32,500 and 16,500 14C yr BP (Willis et Picea pollen have low values even in Picea-dominated al. 2000; Rudner and Sumegi 2001). Finds of pollen of forests and that long-distance transported Picea pollen Quercus, Corylus, Ulmus and Tilia have been attributed do not exceed 5% (Huntley and Birks 1983; Hicks 1994). to their survival in small micro-habitats, in the close prox- For boreal forests it has been demonstrated that Picea imity of the boreal trees mentioned above, which pro- pollen are suppressed by pollen of Pinus (Hicks 1994). vided protection from severe climatic conditions (Willis These lines of evidence give support for a wide spread et al. 2000). The synthesis study by Ravazzi (2002) sug- of Picea-dominated forests between 13,800 and 12,900 gested a relatively wide expansion of Picea during the cal. yr BP in the Gutaiului Mountains and possibly also LGM over central and south-eastern Europe, including in other low and mid- altitude areas in Romania (Fig. 15). the Romanian Carpathians. Charcoal fragments of Picea Larix and Populus pollen have never been identified in sp., Pinus cembra and Salix sp. have also been found in Romanian pollenstratigraphic records, but macro-remains Moldavia (Haessaerts et al. 1998). of these taxa, found in the Gutaiului Mountains (Wohlfarth et al. 2001; Appendix IV), may support the Lateglacial forest development suggestion that Larix and Populus were also present in At the beginning of the Lateglacial interstadial Pinus other parts of Romania during the Lateglacial. Low pol- and Betula started to form open forests at mid-altitude len production and preservation, poor dispersal capac- sites in the Gutailului Mountains, coincident with a de- ity and difficulties in identification usually lead to Larix cline of open vegetation communities. The Pinus-Betula pollen being strongly underrepresented in fossil pollen forests also contained Alnus, Populus and Prunus records (e.g., Lang 1994). The occurrence of pollen and (Wohlfarth et al., 2001; Appendix I, II, IV), while shrubs macrofossils of Larix along with Pinus sylvestris, Pinus (Juniperus and Salix) herbs (predominant Artemisia), cembra, Pinus mugo and few Picea abies and Betula grasses (Poaceae) and sedges (Cyperaceae) dominated sp. was also noted in the eastern Polish Carpathians the open vegetation communities. Our record is in agree- (Ralska-Jasiewiczowa and Latalowa 1996). Larix pollen ment with pollenstratigraphic results from Avrig (Tantau have also been recorded at Kis-Mohos-To in northeast- 2003) and is rather similar to the vegetation pattern re- ern Hungary, but were absent from other constructed at Taul Zanoguti and Iezerul Calimani (Farcas pollenstratigraphic records in Hungary (Willis 1995, 1997, et al. 1999) (Fig. 13). 2000). The presence of Pinus, Picea, Betula, Salix, Around 14,100 cal. yr BP a short-lived re-expansion Juniperus and some thermophilous trees on the Hun- of open vegetation communities and a decline in forest garian plain during the Lateglacial is suggested based components can be observed in the Gutaiului Moun- on pollenstratigraphic records (Willis et al. 1995, 1997). tains. The arrival of Larix decidua is confirmed by macro- Increasing Ulmus pollen values from single grains to remains. This episode, which lasted ca. 300-400 years, around 1% at Preluca Tiganului and around 5% at can also be seen at Avrig, which is situated at a lower Steregoiu were interpreted as a local establishment of elevation (Tantau, 2003), but could not be recognized at Ulmus (Appendix I, II, IV). Since Ulmus pollen have low higher elevation sites (Farcas et al. 1999) (Fig. 14). dispersal capability and the size of the studied sedimen- Marked changes in vegetation occurred at around tary basins is rather small, Ulmus pollen percentages of 13,800 cal. yr BP in the Gutailului Mountains. Trees, which 5% could be explained by its local establishment. This, had already been present, re-expanded, Picea abies ar- however, leads to the questions whether these Ulmus rived and the surrounding slopes became covered by pollen percentages represent small Ulmus populations rather dense forest (Appendix I, II, IV). While the pollen on the surrounding slopes or whether they represent a records show high pollen values for Betula, Pinus and signal from lowland areas or protected valleys (located Alnus and comparatively high percentages for Picea, close to the sites) with favorable local conditions? When the macrofossil records indicate that Picea likely became using pollen analysis alone it is difficult to distinguish the dominant forest taxon (Appendix IV). It is probable between a local versus a regional signal when taxa are that the pollen values for Picea are suppressed by those only present in low numbers. Plant macrofossil analysis of Pinus and Betula, which both are higher pollen-pro- on the same sequence could not confirm the hypothesis ducing taxa and that the macrofossil record more accu- of a local presence of Ulmus (Appendix IV). Similarly, rately illustrates the former forest composition. The for- although few pollen grains of Quercus, Tilia, and est vegetation reconstructed for the Gutaiului Moun- Fraxinus could be recorded in the sequences from the

24 SLOVAKIA 22 24 26 28 UKRAINE Key

Pinus

48 Larix HUNGARY Betula Gutaiului Mts. Iezerul Calimani Salix

Cluj Juniperus 46

46 YUG Avrig

OSL

A VI A Taul Zanogutii Bukarest Black Sea

44

14,100-13,800 cal. yr BP BULGARIA 0 50 km 44 22 24 26 28 30

Figure 14. Map showing the tree taxa, which were present in Romania between 14,100 and 13,800 cal. yr BP, based on pollen and macrofossils (Appendix I, II, I; Wohlfarth et al. 2001) or only on pollen (Farcas et al. 1999; Tantau 2003).

SLOVAKIA 22 24 26 28 UKRAINE Key

Pinus

48 Larix HUNGARY Picea

Gutaiului Mts. Iezerul Calimani Maherus Ulmus Cluj Betula & Alnus 46 Salix

Juniperus 46 YUG Avrig

OSL

A VI A Taul Zanogutii Bukarest Black Sea

44

13,800-12,900 cal. yr BP BULGARIA 0 50 km 44 22 24 26 28 30

Figure 15. Map showing the tree taxa, which were present in Romania between 13,800 and 12,900 cal. yr BP, based on pollen, macrofossil and charcoal evidence (Appendix I, II, IV, Wohlfarth et al. 2001; Lascu 2003) or only on pollen (Farcas et al. 1999; Tantau 2003).

25 SLOVAKIA 22 24 26 28 UKRAINE Key

Pinus 48

HUNGARY Larix

Picea Gutaiului Mts. Iezerul Calimani Cluj Betula & Alnus

Salix 46

Juniperus Mohos

46

YUG Avrig

OSL

A VI A Taul Zanogutii Bukarest Black Sea

44

12,900-11,500 cal. yr BP BULGARIA 0 50 km 44 22 24 26 28 30

Figure 16. Map showing the tree taxa, which were present in Romania between 12,900 and 11,500 cal. yr BP, based on pollen and macrofossil records (Appendix I, II, IV) or only on pollen (Farcas et al. 1999; Tantau et al. 2003; Tantau 2003).

SLOVAKIA 22 24 26 28 UKRAINE Key

Pinus 48

HUNGARY Larix

Picea Gutaiului Mts.

Iezerul Calimani Ulmus Cluj

Betula & Alnus 46 Mohos Salix

46 Hiatus YUG Avrig

OSL

A VI A Taul Zanogutii Bukarest Black Sea

44

11,500-11,250 cal. yr BP BULGARIA 0 50 km 44 22 24 26 28 30

Figure 17. Map showing the tree taxa, which were present in Romania between 11,500 and 11,250 cal. yr BP, based on pollen and macrofossils (Appendix I, II, III, IV) or only pollen (Farcas et al. 1999; Tantau et al. 2003; Tantau 2003).

26 Gutaiului Mountains, no corresponding macrofossils been long-transported from lowlands and these differ- were found to clearly demonstrate their presence during ences could thus just be artificial. The vegetation suc- the Lateglacial. In earlier studies conducted in Romania, cession in the Polish Carpathians is similar to that recon- single pollen grains of these deciduous trees have been structed for the Gutaiului Mountains for this time inter- interpreted as long-distance transported, possibly from val. Betula expanded at the expense of coniferous for- an area located in southern Romania or even further south ests (Ralska-Jasiewiczowa and Latalowa 1996). In the (Farcas et al. 1999; Diaconeasa and Farcas 2002; Tantau Czech Carpathians, Pinus, Juniperus, and Larix were et al. 2003). Pollen of cold-tolerant deciduous trees, such the dominant species during the Lateglacial (Rybnickova as Betula, Salix and Populus in Lateglacial deposits in and Rybnicèk 1996). However, on the Hungarian plain, Europe and North America are often associated with mac- the forest composition seemed to have remained fairly rofossils of these taxa. However, pollen grains of ther- homogenous during the whole Lateglacial interval (Willis mophilous trees have never been confirmed by macro- et al. 1995; Willis 1997). remains (Birks 2003). Based on the composition of today’s boreal forests, Birks (2003) argued that thermophilous Holocene forest development trees could not have been present in a Picea-Larix for- Several external and internal factors, such as competi- est, due to unfavourable climatic conditions. In contrast, tion processes, climate change, migration patterns, soil Willis (1994) and Willis et al. (1995, 1997, 2000) suggested development, anthropogenic impact and physical barri- that thermophilous trees (Ulmus, Quercus, Tilia and ers are generally considered responsible for the Holocene Fraxinus) may have survived in small populations within forest succession (Bennett and Willis 1995). Which of suitable micro-environmental habitats in southeastern these factors played the main role in the Holocene forest and central Europe during the LGM and the Lateglacial. history is, however, still a matter of debate. Below I dis- At 12,900 cal. yr BP the boreal forest started to open cuss the forest dynamics in the Gutaiului Mountains up and from ca. 12,600 cal. yr BP onwards only scattered with the emphasis to evaluate which of theses factors Betula, Larix, Salix and Pinus trees occurred in the may have played the major role in the forest history of Gutaiului Mountains (Fig. 16). Picea pollen percentages Romania. decline and Picea macro remains disappear. Based on The vegetation development at 11,500 cal yr BP, at the good correspondence between the pollen and mac- the onset of the Holocene, was a response to the abrupt rofossil records of Picea during the previous time inter- climate warming reconstructed from many palaeo-ar- vals and during the Holocene, we should expect to find chives around the Northern Hemisphere and was prima- at least some Picea macrofossils, if Picea would have rily triggered by two factors: (i) the Lateglacial forest occurred locally during the Younger Dryas stadial. An composition and (ii) migration processes. At ca. 11,500 explanation for the presence of Picea pollen, but ab- cal yr BP, Betula, Larix and Pinus were the first trees, sence of its macrofossils could be that the local forest which were present and formed open forests in the was strongly depressed during Younger Dryas and that Gutaiului Mountains (Fig. 17). Their immediate response the opening of the vegetation may have facilitated in- to the hemispheric warming implies that they had sur- creased pollen transport and deposition from lowland vived locally during the Younger Dryas. During the first areas or protected valleys, located close to the study 200-300 years of the Holocene, the forests were still open sites. A reduction of arboreal pollen percentages and a and large areas were likely covered by open vegetation maximum development of Betula forests is also inferred communities (Appendix I, II, III, IV). The large number of from the study at Avrig in the lowland (Tantau 2003). At Larix decidua macro remains suggests that a Larix for- the sites Taul Zanogutii, Iezerul Calimani (Farcas et al. est was present at mid-altitudes in the Gutaiului Moun- 1999) and Mohos (Tantau 2003) (with some uncertain- tains and most probably also in other parts of the Roma- ties due to a hiatus), the reduction of pollen percentages nian Carpathians, although Larix has only been noted of Pinus and Betula along with increasing values of non- as isolated pollen grains in other records (Farcas et al. arboreal pollen was interpreted as a distinct decrease in 1999). Picea, Populus, Alnus and Salix were part of these forest cover (Fig. 16). Scattered or low pollen percentage pioneer forests, but their expansion and occurrence was of Picea of around 1% at Avrig, Taul Zanogutii, Iezerul more restricted (see discussion below). Pollen diagrams Calimani and Mohos were attributed to their local pres- from mid-elevation areas in the Carpathians show in- ence (Farcas et al. 1999; Tantau 2003). Overall, the forest creasing values for Pinus and Betula, which was inter- composition shows significant changes in all investi- preted as a Pinus-dominated forest with Betula abun- gated sequences. However, there is some variation be- dant in sheltered areas (Tantau 2003). The beginning of tween sites, such as that Betula became the dominant the Holocene at higher elevation sites was, however, taxon at low and middle altitudes, whereas Pinus domi- characterized by an expansion of Pinus (possibly Pinus nated at higher elevations. Although local climatic con- mugo), Betula and Alnus viridis (Farcas et al. 1999; dition could explain these dissimilarities, the large amount Farcas 2001) (Fig. 17). of Pinus pollen found at high-elevation sites may have

27 At ca. 11,250 cal. yr BP, the landscape became com- sion (Betula, Picea, Ulmus Quercus, Tilia, Fraxinus and pletely covered by forest as new tree species arrived Corylus) was delayed by several hundred years (Ralska- and expanded (Fig. 18). A lower presence of Picea at the Jasiewiczowa and Latalowa 1996; Ralska-Jasiewiczowa onset of the Holocene and its late expansion, some 200- et al. 1998) as compared to our area. This delay may 300 years later as compared to Betula, Pinus and Larix indicate that the trees had to migrate from areas located is seen in all sites in Romania (Farcas et al. 1999; Björkman on the Balkan Peninsula, in Hungary or even along the et al. 2002, 2003; Tantau et al. 2003; Tantau 2003). The Black Sea coast. very similar shape of the pollen percentage curves for The next step in the forest succession was the estab- Picea in Romanian pollen diagrams may indicate that lishment of Quercus, Tilia, Fraxinus and Acer through- Picea started to spread from small local residual popula- out Romania (Fig. 19). Scattered pollen grains of Quercus, tions, located in different areas in the Romanian Tilia and Fraxinus were continuously noted since the Carpathians, which acted as an important repopulation beginning of the Holocene, which may show that these centre for Picea. The delayed expansion of Picea could trees were present in the region. Their local establish- represent the time it needed to build up populations from ment is shown by increasing pollen values, documented local individual trees or, could be due to a migration lag from ca. 10,700 cal. yr BP onwards (~9500 14C yr BP) at from nearby areas, where Picea survived the Younger Steregoiu, Preluca Tiganului, Taul Zanogutii and Avrig. Dryas cold episode. High pollen percentages recorded in the basal sediments Ulmus pollen grains occur sporadically at the onset of sites in the Apuseni Mountains and at Mohos at 10,200 of the Holocene, but increase rapidly from ca. 11,300 cal. cal. yr BP (9000 14C yr BP) show that Quercus, Tilia and yr BP onwards (especially between 10,000 and 9600 cal Fraxinus must have arrived already earlier. In contrast, yr BP), which shows a local establishment of Ulmus at all at Iezerul Calimani these trees were possibly only rarely recently investigated localities in the Romanian present during the whole Holocene period (Farcas et al. Carpathians i.e. Steregoiu, Preluca Tiganului, Iezerul 1999). The establishment of Quercus, Tilia and Fraxinus Calimani, Avrig and Taul Zanogutii (Fig. 18). Although seems to have occurred synchronously, since no or only pollen studies are now available from several key areas a small time difference is observed between sites located in the Romanian Carpathians, no investigations have throughout Romania (Fig. 19). This pattern suggests that yet been carried out in southern and south-eastern Ro- these trees entered Romania from different source areas mania, where potential glacial refugia for thermophilous e.g., from southern, south-western and western Europe. trees may have been located. The establishment of Ulmus If they would only have had one single southern refuge occurred almost synchronously across Romania and the centre, then a distinct, time transgressive south-north slight difference in timing was probably determined by a succession would be seen. Coinciding with their local migration lag from lower elevation sites (where Ulmus establishment, these trees had to compete with already populations survived in suitable microhabitats during established forest constituents, which resulted in the the Lateglacial) to higher elevation sites of each massif. formation of altitudinal forest belts. From this time on- If Ulmus would have migrated from areas situated to the wards, dissimilarities in the forest composition can be south of Romania, then a distinct time-transgressive observed between sites, depending on which of these pattern should be visible during the early Holocene. In- tree species first reached its maximum and on the pro- direct evidence, supporting the hypothesis of a survival portion of tree types already present within the forest of Ulmus in the proximity of the studied sites, could also canopy. For instance, Tilia and Fraxinus reached their be obtained from the fact that Ulmus started to expand maximum 300 years before Quercus in the Gutaiului Moun- 300-500 years earlier as compared to Quercus, Tilia, tains. At Mohos, Fraxinus expanded first and was bet- Fraxinus and Acer. Comparisons to regions around Ro- ter represented. At Taul Zanogutii, Quercus was most mania show a different early Holocene forest succes- important and dominated over Tilia and Fraxinus. In the sion. In Poland and Slovakia Ulmus was the first decidu- Apuseni Mountains Fraxinus pollen percentages were ous tree to expand (Ralska-Jasiewiczowa and Latalowa slightly higher than those of Quercus and Tilia. These 1996; Rybnickova and Rybnicek 1996). In Hungry the differences could be the result of the altitudinal position forest composition varied across the country: Tilia ex- of the studied sites and of local climatic and edaphic panded first in eastern Hungary, Quercus and Corylus conditions, rather than having been caused by different in the northern part and Ulmus and Quercus in the cen- immigration patterns for each forest species. tral part (Willis et al. 1995, 1997). The early Holocene The first regular occurrence of Corylus pollen grains forest of Bulgaria was composed of Ulmus and Quercus is observed in all investigated sites at the beginning of (Bozilova et al. 1996; Bozilova and Tonkov 2000; Tonkov the Holocene. Increasing pollen values for Corylus are et al. 2002; Stefanova and Ammann 2003). Although the noted from 10,300 cal. yr BP (9200 14C yr BP) at Preluca vegetation development in the Polish Carpathians gen- Tiganului, Steregoiu, Avrig and Taul Zanogutii. At ca. erally corresponds to that reconstructed for the Gutaiului 10,190 cal. yr BP (8990 14C yr BP) Corylus pollen was Mountains, the timing of tree establishment and expan- already abundant in the lowermost part of the sediment

28 SLOVAKIA 22 24 26 28 UKRAINE Key Pinus

48 Larix

Picea

HUNGARY Gutaiului Mts. Ulmus Iezerul Calimani Cluj Betula & Alnus

Quercus, Tilia, Fraxinus

Corylus 46 Mohos YUG Avrig

OSL

A VI Taul Zanogutii A Bukarest Black Sea

44 Semenic Mts. 11,250-10,700 cal. yr BP BULGARIA 0 50 km 44 22 24 26 28 30

2000; 2000;pollenFigureMap18. andTantaual. 2003;macrofossil et whichshowing2003).Ttaxa,antauevidence were Romaniatree IV)theinpresentpollen or(Appendix evidence III, betweenonlyII, (Farcas I, 11,250 al.BPandyr 1999;et10,700Röschaccording cal. andFischerto

1999; Bodnariuc1999;pollenFigureMap19.and macrofossilal. 2002; whichetshowing taxa,evidence IV)wereTRomaniapollenantaual. or treeevidence2003;inetpresentthe (Appendix III, 2003).Tonlyantau (RöschbetweenII, I, and10,700 FischerBPand yr2000;9300 Farcascal. based on al., et SLOVAKIA 22 24 26 28 Key UKRAINE Pinus

48 Larix

Picea HUNGARY Gutaiului Mts. Ulmus

Cluj Iezerul Calimani Betula & Alnus

Apuseni Mts. Quercus, Tilia, Fraxinus

Corylus 46 YUG Avrig Mohos

OSL

A VI Taul Zanogutii A Bukarest Black Sea Semenic Mts. 44

10,700 - 9300 cal. yr BP BULGARIA 0 50 km 44 22 24 26 28 30

29 SLOVAKIA 22 24 26 28 UKRAINE Key

Pinus

48 Picea

Ulmus HUNGARY Gutaiului Mts. Betula & Alnus

Cluj Iezerul Calimani Quercus, Tilia, Fraxinus

Corylus Apuseni Mts.

Carpinus 46 YUG Avrig Mohos OSL

A VI A Taul Zanogutii Bukarest Black Sea

44 Semenic Mts. 9300 - 6500 cal. yr BP BULGARIA 0 50 km 44 22 24 26 28 30

antau 2003). Only2003).MapevidenceFigureTantau20. IV recorded(Appendixwhich macrofossilshowingasPicea Farcasistaxa,III, Romaniawere remains ;intree presental. the II, (Appendix1999;etRösch betweenand2000;FischerBodnariuc IV). 9300 BPandyr al65002002; etcal. based. on pollenTantaual., 2003; et

30 OnlyMap evidenceFigure2003).21.recorded macrofossilasFarcasiswhichPicea (Appendix remainsshowingtaxa, Romaniawere III; (Appendixal. intree present 1999;ettheRösch II, and2000;betweenFischerIV).Bodnariuc 6500 BPal2002;andyr4500 et . cal. based Tantaual.on 2003;pollenet, Tantau SLOVAKIA 22 24 26 28 Key UKRAINE Pinus

48 Picea Ulmus HUNGARY Betula & Alnus Gutaiului Mts. Quercus, Tilia, Fraxinus Cluj Iezerul Calimani Corylus

Apuseni Mts. Carpinus

Fagus 46 YUG Avrig Mohos OSL

A VI A Taul Zanogutii Bukarest Black Sea

44 Semenic Mts. 6500 - 4500 cal. yr BP BULGARIA 0 50 km 44 22 24 26 28 30 yr BP(8720 yrsequencesBPCorylus becameonwards(8200from the14 MohosCBP)yr 14 ApuseniCatdominant BP)9600theyrCalimaniIezerul ca. speciesMountains,9800andatyr19).9300from tree(Fig.ca. cal. ca. at allowT2001;beginning2003;antauHoloceneRösch thefromet valuesof . 2000;and sitesBodnariucFischer8500notedareMoreT2003).ca.(Farcasantau BPall yr inal 1999; cal.regularFarcasal2002; et but. et ,. still cal. yr BPyr 8700BPSemenic yrca.7500BPthe yrca. andBPcal. inatthe(8200in(6780atcal. 14 CMountainsBP)(7790PrelucaApuseniyr MohosnorthernTiganului InatHun- BP)20). 14 Mountains,andCat 14 yrandSteregoiu,(Fig.Zanogutiiat aroundyr Taul Bodnariuc9300 7700BPcal. yr ca. BPMountains,III; atlishmentcal. and7500al caexpansionwere2002; occurrences cal. atetlocal(6800ofinterpretedas. however (6700alTCarpinus 2003). antau(Appendix,14 occurred, CBP) et theyr . 14 BP)C inyr TheatApuseni estab- II, Taul , beginning of the Holoceneexpansionthe (Wexpansion componentsmixtum cordedBulgariagary beginningRomaniaofthe(T, ofin of Coylus recordedCorylusal2002;S onkovalis and 1995,et similarfollowedillis ofalreadyand . etdecline2000).isthe Inthatre-to. BPattefanova Quercetum around(5700BPandCalimani.yr Zanogutii(5000asynchroneity at5000 clearSemenicyr14 A theandCBP)14 (4500Prelucacal establishmentinyrtheTiganuluiat inand Avrig Mohos,Mountains,Steregoiuand 670014 CBP)yr5700 Iezerulca. cal. at cal. and its dominance theandAmmann1995-1996;fromforlate(Diaconeasa expansionits western 2001;of FarcasdistantCorylus developmentwas immigrationlocalities THoloceneCorylus wowithwas2003). totheassociatedal2003;andThypothesesantauRomania.havebeenproposed attributedFarcaset in2003) . TheTantau arrival spreadtowardsRomania.graduallyever thenCarpinus expansionSemenicwesternand , seemseasternCarpinus ofRomania.shows Apusenisimultaneouslyhave How-Carpinusdif Mountainsto somea expandedwestevenpatterntheonlyIt expanded canferentbe1000andrecognized.inthus within the infirst late arrival as a consequenceHoweveraas maximum optimum arrivalclaimed. beginningits belatethe2002;climaticSince 2003).Tantau Cor migrationCor (Farcasof expandal.Holocenethe, ofnotoccurreddensity the1999;diddelayylus etofregularly fromFarcas andcan mixtum reachuntilnotQuercetum pollena BPprofiles,all T inantau which Carpinusstudysuchinyr havetablishedfactorsthe yearsforest, includedApusenihamperedstrongwhich , area,aslater competitionwasmany abundantRomanian alreadynotMountains.spreadofmayvery GutaiuluiMountains, infromisthetaxa,farwith Carpinus.Indifferentalready the Aroundes-for cal 6000 - ing forest taxa until for somefor Climaticuntiltaxareasonshad openthatforestforest.diminished.Corylus haveavailableinglikelythe changereplaced inreorganization, gapscausedCorylus forestsomewasFollowingbecameduring thiscouldwhichseemsunablecompeteformer possiblytheto isof thesetwo constituents. forest lines, withexist- the consideredforming isThewaswesternsouth-westernCarpinus Fagus ruptedests,developmentforestswhichexpansionby foreststhebe causedbelt, tolateforestpart. probablydistinct aRomania ofsouthern wasbyin Thelarger increasingandofFagus proportiononlynorthern. north-the ofand orCarpinus inter-thanin cal. yr BPyr UlmusT ecologicalQuer theproportioncal. cidedThe requirements reductionmaximumCoryluswithcus, a offorest theilia, decline ofexpansionslightUlmusa FraxinusofaroundJudgingof thecompared(Appendix its IV).BPand 9300asatfrom yrbased10,300 Corylus coin-on, TFraxinusilia at the, cal. toand III, 1994). immigration Earlierdifferentstudieshaveto Wandpatternexpansion1988).pressurefromillis migrationofRomaniaOtherconnectedspeedroutesintohumanand Fagus from (Behre(Huntleyconsideredfactors refugiachange1988;andCarpinus climatic Birks1983;are theactivities responsible latetheCarpinus Ammann for climatic condition, possible prolongedpossible condition,may(AppendixclimaticlatetheevidenceaquaticsummerCorylus slowexpansionplantmacrofossilclimaticthatTheIV). drierindicateconditions,expansionfor Corylus was seemsinferredmayasinitial fromcommon III, droughtrecord, it ofunder- thehaveby thatdrier beenthe Corylus initiated in gestRomaniarefugiainvestigations(Diaconeasafor theexistenceResultsobtainedhave from(BodnariucCarpinus al2002).anddifferentshelteredproposedinFarcasofglacial1998; refugianeighbouringet placesFarcas2001).More. forwithin existencesug-thecountries Carpinus recentglacialtheonof about 9000 cal. yr BP yr20).andCorylus aboutFraxinusstory 9000dominatedout-competedlayeryrAroundandonwards,oftheexisting forest,cal. only93008000QuerFromca.between latercal. expandedexpand constituents.atBPyrtoT Picea Ulmus,startedforestCorylus these cal.rapidly,9000 BP yrandcus, sig- (W6500(Fig. ilia and cal. However 2000).and(WCarpinus ment westernpandedBalkan(Bozilovagary Europe.general alof southernPeninsulaillis thelocaland(Figs.etforestinEuropeT20,1994), onkov21),Atwhicheastern ., 2000; BulgariamyinorHun-aland T2002)onkovillis or ex-earlythe establish-, Carpinus et study. first in sites, Fagus nificantly andbecame expansionpresentthefollowsCorylus nificantly acrossveryRomaniasuccessionSporadicpollensimilardeclineof intheco-dominantof diagramsis Carpinus and mixtumQuercetum with Fagus 21).andpollenThesince(Fig.severaland precedesofgrainstheare sites Corylus . al1999; RöschwasFagus frometobtained2003).pollenBPbyfollowedThevalues. severalRomaniayrconsecutivefrom 8400in occurrencesonwards,and scatteredinvestigationsotherFagus, Fischercal. 2000;nicelysamplescorresponds ofFarcasfollowed 2001; (Farcas byFagus increasingpollengrains results31 to T antau SLOVAKIA 22 24 26 28 UKRAINE Key Picea

48 Ulmus

Betula & Alnus HUNGARY Gutaiului Mts. Quercus, Tilia, Fraxinus

Iezerul Calimani Corylus

Carpinus Apuseni Mts. Fagus

46 Avrig Mohos YUG

OSL Taul Zanogutii A VI A Bukarest Black Sea Semenic Mts. 44

4500 - 3500 cal. yr BP BULGARIA 0 50 km 44 22 24 26 28 30

2003). OnlyMap evidenceFigure2003).22.recorded macrofossilwhichasPicea Farcasis(Appendix showing remainstaxa, RomaniawereIII; (Appendixal.intree present1999; theetRösch II, betweenand2000;IV).FischerBodnariuc 4500 andBPal 2002;35000yr et . cal. basedTantaual. on2003;pollenet, Tantau

32 2003 evidenceFigureMap23.). Farcas(Appendixwhich showing taxa,III; al.were Romania1999;treeet inRöschthepresent II, and2000;FischerBodnariuc between BPal 35002002;yr et -300 . cal. basedTantaual. on2003;pollenet, Tantau SLOVAKIA 22 24 26 28 Key UKRAINE Pinus

Picea 48 Ulmus

HUNGARY Betula & Alnus Gutaiului Mts. Quercus, Tilia, Fraxinus Iezerul Calimani

Carpinus

Apuseni Mts. Fagus

46 Abies Avrig Mohos YUG

OSL

A VI Taul Zanogutii A Bukarest Black Sea

44 Semenic Mts.

3500 - 300 cal. yr BP BULGARIA 0 50 km 44 22 24 26 28 30 SLOVAKIA 22 24 26 28 UKRAINE Key Picea Pinus

Ulmus 48

Betula & Alnus HUNGARY Quercus, Tilia, Fraxinus Gutaiului Mts. Corylus Iezerul Calimani

Carpinus Apuseni Mts. Fagus

46 Abies YUGO Avrig Mohos

SLAV

IA Taul Zanogutii Bukarest Black Sea

44 Semenic Mts.

last 300 years BULGARIA 0 50 km 44 22 24 26 28 30

(AppendixFigureFarcasMap24. III; al. showing1999; etRöschII, which taxa, and 2000;treeFischerwereBodnariucthe Romaniapresentin al 2002;300 et based . years,during last ontheTpollenantaual. evidence2003; et 2003).Tantau cal. yr BPyr maniathefromcal. presence(around dominant , 22-24).about6600Fagus showsTheyears(Figs.BP),the4000 localBPinvestigatedinforest(Appendixyrmay highestcal. presentstrongproportion theRo-allincreasingina4800andcal.totreeforest trendpastsites of becamebean evidenceFagus ca. III). Atwhere its forII, (Rösch BodnariucMountains,2001;(Appendixregion,sites, Gutaiuluiriod fromalRomanianand but2002;whichFischer ThisIII) othernotAbies 2000;was. theetincloseaparticular Farcasdiffers . iswellvicinity representedsituationalTof the1999;antaual.study Farcas2003). foretthepe-during . sitesthis lated to a changea mid towasclearance Mountains,2000;GutaiuluilatedandniaatpresentbytowardshumansFischerconnected elevations.TheFagus expansion forestmorewoodlandsalthougha humantolateal Tantau2003), humidRoma-(Röschwas et ofwhile. was evidencedinfluencecorre-it situatednotin theFagus isclimate in sites at arborealreductionindicationatenTheHumanin pollenanthropogenicof density percentagesforestof first study areaevidenced inalong non- , disturbancein withdeclinebyindicators arise of-a is files withinRomania dominance withincidingrecorded. files cantcidingexcludedthe declinewithdecline diversityThis(Appendixforestainalso theseenfeature22). BPofisotherFagusexceptvalues(Fig.have inAroundyrpollen been Quercus III). peaks3500, pollen forpro- forest,asignifi-co- cal. ca. coin-Quer cus is Cyperaceae,bypercentagesGutaiului1988;arborealpollentypes(IversenRanunculaceae,1949,increasingMountains Ammann4000Ber1960; valuesmainlyduringwetland oflast Behreyearsthe is non-arborealcaused 1989; increaseglundinspeciesFilipendula1981,the such2000,pollen as the2003).and In Potentilla . forest densityMywhenSwiss tifully Fagus (Ammannfoundmayforestantiphasesthe . diagramsinpollen duringHolocene,show Holocenedobe nothigherpossiblelate theAlps thatpollenlate Quercus anyduringfloweredvaluesthemore although for1989) ofplen-and- were reductionQuer Quercu byexplainedit a cus s suchgoodhighareentsmentChangessignificantvaluesas pollen valuesfor Betula,3500existingPinus,indicatorsofBPtaxa ca.clearance.yrforestAlnus,byof Fagus Carpinus Picea, compositionatandthepioneerstudycal.Inin andreplace-area constitu-forestQuercus coincideand Corylus (i.e. foresttheof ) 300 cal. yr BPyr markedThisthena24).300Quercus ratherarerelatedincreasedalthoughto thecal. Abiesrecordedinmayforest.between are pollenregularlyvery, taxoninfloweringlow this3000increaseindicatewasproportionthepercentagesthatin denseandless ainpresent 23,theofin(Fig. forest, successionpollenmostconditions,forestwith percentageshumansincestrongpercentages.a declineofT drivenFraxinusabsentfor Thisby(AppendixandchangeBetula Ulmus, wasaThedisturbance ilia, climaticincreasewaswereAlnus naturalinterpretedindicatorsparalleldrier aandasIII). intointerpreted33 al- Acer II, the goodspreadingapeattheas surface correspondence (Appendix On thesearoundoftrees andbetweenothertheIII). hand, sitetheover spreadingtheofII, 300 grazingreductionlast yearshuman and diversityforestinpastureland. and reflectedbyDuring antwoisabrupt increaseactivity centuries,forestlast thein to forest clearancesuch reductionforest thethe tocies,ousFagus disappearance andasduring300 Alnustreesforestpreviously, re-expansionof theregeneration. lasttheyears Betula, Thecoincidesestablishedand opportunisticthedecidu- ofalmostreductionwith otherspecies,Pinus couldspe- ex-treepoint ofalsoall purpose(Röschdiagramsevidencetransformedingwood 2000;andBodnariucfromformerlyFisherthehuman Romanialowlandwasforforested6000-7000exploitationopen into pastseenundertaken wereseveral areastheal.inpollen issettlementindustryand2002;etyearsclearfor heat-contrast,In activities areas. T antau dle of the last century (Karocsanyilast the offordlereforestationobserveddays,manceptbyenclaveswhichplantation, Quercus withinthe animpactlandscapeNowa-argument1995)., on(Appendixconifers,IV).theofespeciallyFagus istookforest.This mid-placetheofstrongthe inahu- resultbecanisII, offor Picea , pollencultivatedandthebasinscouldThe2003),extensionsourcecanopy closedmainlyonlyaanthropogenicratherInweakfields, forestlateaalso bearea.i.e. due forested ofgrazingthesmalllandscape humanto theinvestigatedofwhichsizeinduced areasandclearance. studyreducesareatheforest, infires signal g. majorP Asteraceaeacetosa/R.g. Behreceous PlantagoAnotherhumanmediaacetosella,taxaTubulifloraeforindicating 1981,sign , somegrazing. , Chenopodiaceae,Asteraceae 1988).areinfluencesRumex , andCannabaceae Pollenherba- grazingLiguliflorae, forestindicators Artemisia (e.suchof, Ur (e.g., tica 1991; activities , pollenBroström,representingconsiderableaopenal realmand past. sedimentary2002). underestimationaislandscapes there (Moorthannon-arbo-1000basin withlessradiusof a et ties. From wildgrowingincreaseslightorBP (afromfenyrHolocene2300activitiesties. deriveSteregoiuginningCentaurea –thecyanuspolleninnaturalherbcommuni-ofon andmorecal. pollenhumandiagrams documentedpercentagesthePrelucaof signsare type)inTiganului, clearfrom thesemightbut be-alsoforthe since ofthevegetation andchangessitivityThisbyLateglacialClimatic chapterHolocenefluctuationsvegetationandresponsecontains discussionthe15,000duringa its focusedlast theinclimateGutaiuluitoyears inonas reflectedsen-the forest grazing and300forestacetosella,Chenopodiaceae,Plantagoanexpansion asmaythearea yearsgrazinginlanceolata, around)andpressure becameopenforestPoaceaeof Cannabis- >40Rumexbe landscapesDuringinterpreted antheasstudied introductiontype , site. ofµacetosa/R. m, Artemisia thelast , GRIPRomaniarecordsthelakeandcomparedMountains. incaves) to corefromwesternThe iceEuropeantoGreenland.andvegetationchanges alsorecords furthermoreto fromare (lakesotherinvestigated sites mires is dominated mentionedandmiressivelymcatorsmoreexploitedmedia/majoris Plantago bypasturePoaceaehasasanopen Theland. intense,belowinferredasabove,(mainly occurrencecharacterfromandtheplusvegetation . T800openingsodayforest thepollenofthe ofriseindi-and inten- Molinia are BPUrtica plantcommunitiesmacrofossils, yr conditions(Appendixlowclimaticetationsediments>14,700Terrestrial IV). calbasalthe(Appendixminerogenicin productivityextensiveand alsoinferredindicateimplyfrom ColdsteppeIV). andlowis temperatureswatersub-alpinelakeaquatictheI, withveg-andtaxadry lanceolata derivefewtroductiondomesticGlyceria andfromwere Cyperaceae,pollenofgrains)>40mm Poaceaeanimals. found,cereals),Humanof cultivatedtheyprobablybutSecale Cannabis- mayfromandwhich, plants,(whichparticularlyoriginate type,usedfeed alsocereals. toareseen in-by Plantago alsotheimpact Ais andvariousEurope seenBirkscentralGI-1e)LateglacialTheBjörckrecordsin (14,700from1 is–Ammann rapidwarmingnorth-western(e.g., BP) al. yr1998;et and 1,50014,700cal ca.Wofal.BPalkerLoweyr1999;trendat 2000;etcal Renssen(start Isarin 2000;and point to human-inducedHowever tocharcoalHolocene,inpointgatedpressioncultivated concentrationslandlowersites. areasandAnwhichat Oasthefires. innotwasaltitudesfrom wouldoccurrenceincreaseDe-charcoalmajortheof nocouldnotedagriculture particlesindicateparticularalsoduring, closeinvesti-the risetolatethe Gutaiului2001;AlthoughthepollenmentThismineraland thelithology V andenber percentagesmagneticMountainspollentemperaturemacrofossil almost increaseal., ghe imme-tree 2001;etparametersassemblages 2001;Hoekclearly registered 25)Wis(Appendixalkerand(Fig.risesedi-IV).2001).in the twothe of II, sites I, In summaryBPyr34 forestexpansionaroundvegetationhumanfires.signsofIn the2300 mostandarecal. mainly, pastureareof landandgrazing. , evidentDuringrepresented pollenimpactforestthe instratigraphybyan onlocal the 14,700gradualdevelopment BPaatyrvegetationforesttimeeredingpointsdiatelytoimpliedSinceslopes.calbylag tree-lessbeforesurroundingsby thepollen 14,700andmacrofossilplantmacrofossilsthe plantwereBP yron surround-cal, mostthesite recordofcov- the , vived during the Last Glacial MaximumGlacialLastvivedprobably duringthe representsmigrationa Trees, whichlag. lowlandsur- inareas, between1998). al. stratigraphy possible13,150A explanationand 12,900 BPmightyr thisfor(Björck bethat et stands or increasing ofperaturesstandscreasearborealgraduallyscatteredresponsepollendensity inpercentagestrees forestand productivity, colonizedcanhigher Similartomaylocallyexistingbeofonly openregarded .tem- latertrendshigher withreflectionthusrepresenta asinelevations.anpredominantlysignifi- in-theriseThe that recordingisinitial temperatureRomaniainlagtionwhichernthe BPA forestedyrlandscapedecreasesuch 13,800beganre-expansion wasshortclimatic hamperdevelopment,cal. mayatcoolingopen notandonlyminorofdid northwest- forest events.explana-vegetation. An inthewithstandsof alternative the sequencechangesopenCalimaniof fromsiontionandthecantlyAincreasing Iezerulvegetation300markedseemcommunities, whereasca.(Farcasyear valueslonghave interval, Avrig, arboreal toofal.onlyobservedoccurredminor 1999;byre-expan-pollenet ais atseen vegeta-areZanogutii2003).be- Tinantau aul seemsBPhaveyr(Appendixbutbecame to cipitationevidencecal.on,macrofossilBetula observedinferredfrom(as , stronglyLarix is IV)moredecliningfinds)for , andtemperaturesSalix Thisfrompollen).300pronounced theI, 12,900changeBPfrom yrinferredca.Pinus 12,600year (asandtimean cal.be-lagpre- at is (Appendixshorttweenelevation(T pollen(Farcasepisodeantaupercentages2003), IV).whereassites cooler of14,100phase,similarandhasal.beenAwithalso I, climaticnoet1999). lowerchangesdrierconditionsandTheobserved 13,800BParborealrecordedhigheryrresponse areathe atasinterpretedofcal. and Ais vrig speleothemsvegetationplainedtweenare2003).alsoindicationsIn bycommunities, response,pollenregionalcoolerandforathefromwithpollen macrofossilre-occurrenceBihorand thesource. mayMountainsclimaticrecordedsignalsdrier openofcondi- (Tis atAnbeex-strongthere Aanalogous,vrig antau nia, whichtime Gutaiuluivegetationingnia,speleothemsconsistentcorrespond(TthistheintervalNWMountainsin coolingimplyfromwiththe phasethe toMountains, recent (GI-1d)coolerApuseniamasstableand isotopeclimaticconditions2003).recognized drieronand spanresultstimecouldThisdur- Roma- inis A vrig(Twastime Romania.intervalmostrevealed thechangesepisodetogetherwhichtions signal this isin pronounced amas at2003).theLateglacial,of higherareZanogutiiprominent,less(FarcasleadingTAaul elevationApparently vegetationresponsetheandCalimani,distinctsites toIezerulal,1999).cold distincttheof et . but still Al- nant forest componentsoon13,200 BPyrre-expanded forest followednantattains, and NorthexpansionBP calAroundAlnus forestsbyrecordsAtlanticGutaiuluiyr(Björckmixed themayMoun- andby inestablishmentits the 13,800 alcal. Ulmus 1998).have . beenbecamedomi-Picea et of Betula . the , Pinus and, Picea , Larix GS-1coldand recordedormatelyof800vegetationclimatic(Björckcoolingthe aroundwasNorthdrytomal.climaticwithbeginningconditionsthe 1998).et weakeramplitudea.s.l. strongestcoincides belowtheof AtlanticAlthoughapproxi-region, higherits Y oungerelevation.atTheclearly Dryasperiod isstart stadial this ture content fromcontent BP(T aroundyrat agespendixturefavouredZanogutiiIV).TBPaul A13,800yrmarked cal(Farcasvrig temperatures byonwardscal 2003)antaurisingal and. ofThe1999)and etresponsearborealmois-increase(Ap-pollen increasinginpercent-wasvalues pollenPicea soildistinct thenotedof alsoat However(WEurope.of2000).partchanges hashigh been show2000;maticattalwasPeteetfrom1994;It elevation shownRomaniaprobablymore areasresults rapidand pronouncedaistheillis more lessvegetationresponse the, Ammannsensitivealcontinen- minorthat2000;vegeta-to thedistincteastern,of theet in. cli- Wick ated at lowmust supportedThisatlimit vegetationis atedforesthighertains.indicatedtreebymid- haveGutaiuluimacrofossil theRomania,elevationby ,Moun-in beenandsignalthefindsremarkable. encounteredareas (1800highwhichlocatedis abovealtitudes,time mthisaswhichat 800situ-a.s.l.), locatedseensitesThe monly belowsitesatis aroundencountered atisweaka.s.l., maynalencetions,onlytiona response in10001800whereasstronglowaresponse inWhydepressionreflectpattern? andwasmThemid-altitude coolingandtheobservedlimit nothata.s.l., areashighera.s.l. thetreethereofsucheleva-orfacta sig- atdiffer-isa 12,900 cal. yr BPOverall,yrbetween intervalGI-1c,thelimit. landthe 12,900forestprobablymay 1binterstadialAllerødwhichcal. andmajorpollenzone Scandinaviabecameregarded13,800thebe1aandwarmcorresponds, ofca.(Björck longestas Green-approximatelytheandwereor toperiodforested, al.thecloselytolocated1998), et BetulaThemaybyAlnus,foundlowering highhigher(moreelevationslowercould at sitesloworiginatingaltitude ecotone.pollen)thefrompronouncedberegionalofproduction aexplainedsignal, Therefore,pollen lowland representation pollenandthesignal (otherthetransportcapacity offortreesareas. (e.g. Picea, Pinus ) during the Lateglacial in Romania. in Lateglacialsedimentseventcoolingnianduring1996;howeverGI-1b,pollenstratigraphiesrecords,the orEuropeGRIPNoneAmmannalwhich from2000)phase,the Roma-central (Brooksandsuchet in so-called. hastheofbeenas intra-Allerød short-termalrecognised1997;acold lake et ice-corein, . Wick butcouldognitionlowOnpercentagesPicea of), chronologicalmayotherhand,thesealsocontrol ofthebechanges low-resolutionPinus suppressduehavehigh higherthatthe facthamperedattopollenanalysisandelevation.those sites othertheinoftaxa. rec- 35 the tal grey 36 thesedimentbars inrepresentfrom talinferredFigureassemblagesSummary 25. hiatii Preluca sequence.Tiganului andLateglacialHoloceneand thezones.of Steregoiu,PForchartbased MAZdetailedadescription environmentalonlithologymacrofossils,theofpeat;=pollen,seelithology and sedimentdevelopmentT ablesandclimatic1assemblagesmacrofossil=in2 IVparameters.LPGutaiuluizones; MountainstheAppendix The in . horizon- AZpollen local = Holocene documented by oxygen isotope records from northwest- Distinct climatic oscillations during the Holocene were ern Romania (Tamas 2003) and show the same pattern as recorded in speleothems in northwestern Romania and in the North Atlantic region. In contrast, isotope records are well correlated to those evidenced in the North At- on stalagmites from southwestern Romania indicate a lantic region (Onac et al. 2002; Tamas 2003). In contrast gradual temperature rise and a climatic optimum much such distinct climate shifts are hard to distinguish based later, at around 5000 cal. yr BP (Constantin 2003). on pollenstratigraphy and inferred vegetation changes The studies in the Gutaiului Mountains allow detect- (Appendix III and IV). Pollenstratigraphic records for ing some changes in the forest structure during the early Romania were usually analysed with respect to the time and middle Holocene, which could not have been re- of arrival, expansion and regression of Holocene forest lated to migration processes (Fig. 25). At ca. 10,300 cal. taxa and less attention was given to climate as a driving yr BP, Ulmus declined permanently and Corylus ex- force for forest dynamics. A lower resolution of the pol- panded and achieved dominance around 9300 cal. yr BP. len stratigraphy and sometimes poor dating control may This was interpreted to be climatically induced, i.e. con- have prevented the recognition of the short-lived cool- nected with a dry and possibly cold period (Appendix ing phases seen in speleothem records. III, IV). These climatic conditions may have affected the The diversity and abundance of macrofossils and forest structure and density and the gaps left by Ulmus pollen of aquatic plants increased around 11,800 cal. yr were then occupied by Corylus. Two isolated charcoal BP, coinciding approximately with the later part of the peaks are noted at 9030 and 8700 cal. yr BP, but these Younger Dryas stadial (Appendix IV). More pronounced could indicate local fires. In Appendix III and IV, the veg- changes in aquatic pollen and macrofossils are observed etation changes around 10,200 and 9300 cal. yr BP, re- at 11,500 cal. yr BP, simultaneously with the response of spectively were linked to two distinct climatic cooling the terrestrial vegetation to the significant warming at events centred around 10,300 and 9500 cal. yr BP in North the beginning of the Holocene. The macrofossil and Atlantic records (Bond et al. 1997, 2001; Björck et al. pollen records from Preluca Tiganului indicate moist con- 1998, 2001; von Grafenstein et al. 1999; Nesje et al. 2001; ditions already at 11,500 cal. yr BP. The rise in tempera- Seppä et al. 2002; Hannon et al. 2003). Shorter, colder ture and probably also in moisture availability were the phases during the early and middle Holocene have also triggering mechanisms, which initiated the vegetation been registered in speleothems from northwestern and succession, development and migration at the start of southwestern Romania (Tamas 2003; Constantin 2003). the Holocene (Appendix I, II, III, IV). Although Betula Increasing forest fires and a forest reorganization re- and Pinus expanded rapidly, a time lag of ca. 300 years as corded between 10,200 and 9650 cal. yr BP may also be a compared to the response of the aquatics is observed. consequence of dry climatic conditions (Bodnariuc et This time lag most probably represents the period needed al. 2002; pers. comm.). These climate oscillations com- by these trees to enlarge their populations and to de- pare well to changes in the forest composition and to the velop extensive forests (Figs. 17, 25). A rapid and fairly climatic reconstructions obtained in the Gutaiului Moun- similar response of the vegetation to the rapid warming tains (Appendix IV). A distinct decline in Quercus, Tilia, at the start of the Holocene is reconstructed from other Fraxinus and a re-expansion of Picea over the peat sur- pollen diagrams in Romania (Farcas et al. 1999; Tantau face and on the surroundings slopes between 8600 and et al. 2003; Tantau 2003). 8200 cal. yr BP coincides with maximum dry conditions The abrupt response of Ulmus (200 years after the (Appendix III, IV). A cool and dry episode centred around start of the Holocene) implies a distinct temperature and 8000 cal. yr BP is seen in stable isotope analyses of moisture increase (Fig. 25). The expansion and domi- speleothems in the Apuseni Mountains and in south- nance of Quercus, Tilia, Ulmus, and Fraxinus between western Romania (Constantin 2003; Tamas 2003). All 10,700 and 8600 cal. yr BP may point to a more continen- these changes have been connected to the 8200 cal. yr tal climate with higher summer temperatures and lower BP climatic reversal (Appendix III, IV; Constantin 2003; precipitation as compared to the present (Appendix III, Tamas 2003), which is documented by several proxies IV). Furthermore, the remarkable high concentration of (including pollen and macrofossil records) in the North micro-charcoal between 10,600 and 10,400 cal. yr BP could Atlantic region (Nesje and Dahl 2001; Seppä and Birks be connected to increased fire frequencies, which may 2002; Baldini et al. 2002), in central Europe (von have been the result of drier summers. The subsequent, Grafenstein et al. 1998; Tinner and Lotter 2001; Wagner continuous decrease in moisture from about 10,300 cal. et al. 2002; Magny et al. 2003; Ralska-Jasiewiczowa et yr BP onwards is documented in the aquatic plant record al. 1998), in southern Europe (Wick et al. 2003) and in (Appendix IV). High temperatures between 10,750 and the Mediterranean Sea (Ariztegui et al. 2000). In con- 8600 cal. yr BP may have triggered the spread of thermo- trast, other authors in Romania have attributed the maxi- philous forests to their widest and highest distribution mum occurrence of Corylus to the Holocene thermal op- during the Holocene. Higher temperatures during the timum (8000-5000 14C yr BP or ca. 9000-5000 cal. yr BP), first part of the Holocene (11,500-8200 cal. yr BP) are which was the warmest and most humid interval during

37 the Holocene in Romania (Diaconeasa and Farcas 1995- lution pollenstratigraphy, complemented by stud- 1996, 2002; Farcas et al. 1999; Tantau et al. 2003; Tantau ies of macrofossil remains, mineral magnetic param- 2003). eters and organic matter analyses, as well as AMS The pronounced Picea recurrence around 8200 cal. 14C measurements. The combination of yr BP (confirmed by abundant macrofossils) is contem- pollenstratigraphy and plant macrofossil analysis poraneous with a slight re-expansion of mesophilous was important for the interpretation of vegetation trees (Appendix III, IV) and low frequencies of charcoal changes in terms of local presence/absence of some (Fig. 25). This and the association of Corylus and Picea trees during the Lateglacial. between 8000 and 5700 cal. yr BP could imply less natu- ral fires, lower summer temperatures and moister climate. § During the Lateglacial period vegetation dynamics Geographically contrasting climatic conditions are shown were likely driven by climatic fluctuations and by by oxygen isotopes from stalagmites between 8200 and physiological requirements and tolerances of the 5000 cal. yr BP: stalagmite records from northwestern different taxa. Prior to 14,700 cal. yr BP, open veg- Romania indicate a cool and possible dry climate (Tamas, etation communities and possibly scattered stands 2003) or, similar conditions as at present (Onac et al. of Pinus and Betula colonized the surrounding land- 2002), whereas higher temperatures are inferred for south- scape. Unstable slopes, which were highly suscep- west Romania (Constantin 2003). These reconstructions tible to erosion, may have existed around the basin compare well with present climatic conditions, i.e. the as indicated by the minerogenic bottom sediments. south-western part is influenced by warm air masses from Open Pinus-Betula forests and more stable soil con- the Sub-Mediterranean region, while the north-western ditions developed between 14,700 and 14,100 cal. yr part receives oceanic air masses from the North Atlantic. BP and were followed by a return to open vegeta- Altogether, pollen and stable isotope data indicate that tion with stands of Pinus and Betula. Around 13,800 the north-western part of Romania may have experienced cal. yr BP Picea became established and soon domi- more humid conditions between 8000 and 5000 cal yr BP, nated the landscape, together with Pinus, Betula, as compared to the south-western part. The expansion Alnus and Larix. Ulmus may have existed locally in of Carpinus at ca. 5500 cal. yr BP, which coincided with small populations. Forest fires may have been com- a regression of Picea and Corylus and a rise of Tilia and mon between 13,400 and 12,500 cal. yr BP, possibly Quercus, may have been favored by a change to warmer indicating dry climatic conditions. Open vegetation and drier climatic conditions (Appendix III). This is sup- with stands of Pinus, Betula, Larix, Alnus and Salix ported by high pollen concentrations, low peat accumu- occurred between 12,900 and 11,500 cal. yr BP. The lation rate and a reduction of wetland indicators (Ap- phases with open vegetation were correlated to pendix III). Also the increase of micro-charcoal particles events GS-2, GI-1d and GS-1 and the phases with could indicate drier conditions. The expansion of Fagus tree expansion to event GI-1e and GI-1a to c of the at 4800 cal. yr BP may have been related to possibly GRIP event stratigraphy. If this correlation is cor- higher humidity, i.e. milder winters, and to more acidic rect, it shows that Lateglacial climatic variations were soils, although human activities cannot be ruled out (Ap- driving factors for the vegetation dynamics in north- pendix III). The dominance of Fagus from 4000 cal. yr BP western Romania. onwards was most likely climatically induced, favoured by more humid and possibly cooler summers. The present § Although distinct changes in forest density can be distribution of the Fagus forest in Romania is related to recognized at 11,500 cal. yr BP, i.e. at the Lateglacial/ a moist and cooler climate. A dry period may have oc- Holocene transition, the aquatic macrofossil record curred around 3500 cal. yr BP, indicated by a strong re- gives indications for increasing temperatures already expansion of Quercus on the slopes surrounding the at 11,800 cal. yr BP. basins, a reduction of wetland taxa, by exceptionally high pollen concentrations (Appendix III) and an increase in § Forest dynamics during the Holocene were driven charcoal particles. A warm and dry period is also docu- by complex processes probably caused by the in- mented by speleothems records in both north-western teraction of several factors, such as forest succes- and south-western Romania (Onac et al. 2002, Constantin sion, immigration lag, inter-specific competition, 2003), which implies a broad amplitude of this event. changing climatic conditions, soils and - during the late Holocene also human influence. The vegeta- tion response at 11,500 cal. yr BP (expansion of Conclusions Betula, Pinus, Larix and Alnus) can be interpreted as a succession process, because these trees were § The vegetation and climate development during the already present in the area. The expansion of Picea past 15,000 years in the Gutaiului Mountains of north- and Ulmus at 11,250 cal. yr BP was likely due to a western Romania has been inferred from high-reso- migration process from closely located residual

38 populations and the later expansion of Quercus, ducing me to pollen techniques and the Tilia program, Tilia, Fraxinus and Corylus at 10,700 cal. yr BP correcting my manuscripts, and advice while I was in shows a migration from more distant localities. As Romania and Stockholm. I am profoundly grateful to Ann- the forest became denser, competition between new Marie Robertsson, my co-supervisor, for her patience, arriving taxa and already existing forest constitu- kindness, and suggestions for improvements of the ents played an important role in the forest structure manuscripts, especially the synthesis of the thesis. and may have been triggered by changing climatic Thanks are also due to Prof. Iustinian Petrescu my su- conditions. It is speculated that the late establish- pervisor at the Department of Geology, Cluj University, ment of Carpinus at ca. 5700 cal. yr BP and of Fagus Romania. at ca. 5200 cal. yr BP could have been caused by Thanks to Bogdan Onac for including me in this in- climatic changes. Human influence can, however, teresting research project, for help during the fieldwork, not be excluded for a delayed expansion of Fagus, for always being positive about my work, and giving although evidences for this are not clear. A distinct stimulating comments on the manuscripts. human influence is evident in the pollen stratigraphy Thomas Persson and Sven Karlsson gave kindly as- around 2300 cal. yr BP and is related to the exten- sistance with the tricky and sometimes annoying Tilia sion of pasture land and forest grazing in proximity program. Mats Rundgren introduced me to pollen prepa- to the sites. During the last 300 years human activ- ration techniques. I thank Kajsa Cinthio, Per Sandgren, ity is reflected by an abrupt reduction in forest di- and Oana Stan for helping me with magnetic measure- versity, and an increase in forest grazing and ments. I am thankful to Tudor Tamas for his constructive pastureland. criticism on the manuscripts and for the enormous help with the drawing program to produce the maps. I would § The present study gives evidence for the existence like to extend my thanks to Siwan Davies and Joe Kearns of the following trees during the Lateglacial at a for correcting the language. I am greatful to Helena mid-altitude area in the Gutaiului Mountains: Pinus, Alexanderson for her kindly help with the layout of the Betula, Picea, Larix, Populus, Salix and Alnus. Sig- thesis. nificant pollen percentages for Ulmus indicate that I would also like to thank all my colleagues and PhD this tree could have been present in this area be- students at the Department of Physical Geography and tween 13,200 and 12,900 cal. yr BP. However, iso- Quaternary Geology, Stockholm University for the stimu- lated pollen grains of Quercus, Tilia, Fraxinus and lating environment and their friendship. Also, I give Corylus are attributed to long-distance transport. thanks to the following friends who made my time easier while I have been in Stockholm: Siwan, Dani, Teodora, § Coniferous and cold-tolerant deciduous trees have Amélie, Yoshi, and Veronica. probably been present in Romania already before The Stockholm University financed my doctoral po- 14,700 cal. yr BP and may have survived the Last sition at the Department of Physical Geography and Glacial Maximum. Since Ulmus seems to have ex- Quaternary Geology. Fieldwork in Romania was financed panded rapidly in Romania shortly after the begin- by the Swedish Research Council and Kungliga ning of the Holocene, it is assumed that it at least Fysiografiska Sällskapeti Lund (grants to Barbara survived during the Younger Dryas stadial. It re- Wohlfarth and Leif Björkman). Financial support for mains, however, uncertain whether Ulmus was scholarship was also provided from the Erasmus pro- present before 13,200 cal. yr BP. No evidence exists gram, the Swedish Institute and the Royal Swedish Acad- so far from Romania to prove the survival of Quercus, emy of Science. Tilia, Fraxinus and Corylus before and during the I am thankful to all my dear friends, especially to Lateglacial. Sanda, for their moral support and good thoughts. My warmest thoughts and thanks to my family (mama, tata, Renata, Ovidiu) and in-laws ( Lucian, Adrian, mother Acknowledgments and father inlaws) for their enormous support, for al- ways caring and being there for me. I am thankful to First, I would like to thank my head supervisor, Bar- Markus for being a special part of my life. bara Wohlarth, who initiated the project and gave me the opportunity to work in this fascinating field of science. I am very appreciative for her enormous support and guid- Svensk sammanfattning ance in the Quaternary world, of which I had little knowl- edge before I started this work. In addition, I would like Klimat- och vegetationsförändringar under de senaste to thank Barbara for her hospitality while I was visiting 15 000 åren är väl kända från området kring Nordatlan- Lund University. I am very grateful to my co-supervisor ten, medan enbart ett fåtal studier belyser utvecklingen i in Lund, Leif Björkman, for his help and patience in intro- sydöstra Europa i detalj. Rumänien har en lång tradition

39 när det gäller pollenstratigrafiska undersökningar och En jämförelse med de få pollenstratigrafiska studier rekonstruktion av skogs- och vegetationsutvecklingen, som gjorts i Rumänien under senare år visar att men kronologin för de förändringarna som skedde är vegetationsutvecklingen i Gutaiuluibergen var likartad i ännu mycket osäker. Många av de äldre undersökning- låglandsområden, medan högre liggande lokaler inte arna är översiktliga och uppfyller därför inte kraven på uppvisar lika stora förändringar under slutfasen av sista moderna paleoekologiska undersökningar. Huvudsyftet istiden. Trädslagens invandring och spridning under de med avhandlingen är därför att senaste 11 500 åren verkar ha skett samtidigt över stora · upprätta en mycket detaljerad tidsskala för de delar av Rumänien. Däremot skiljer sig den låga repre- klimat- och miljöförändringar som skett under sentationen av avenbok och den svaga mänskliga på- de senaste 15 000 åren i nordvästra Rumänien, verkan inom det undersökta området tydligt från andra · klarlägga skogens historia områden. · diskutera de processer som kan ha styrt och Under de kallaste faserna av sista istiden har björk, påverkat skogens dynamiska förändringar. asp, tall och gran funnits i låglandsområden i Rumänien. Undersökningarna baseras på sedimentkärnor från två Alm växte troligen i mer skyddade områden (refugier) i små före detta kratersjöar, Preluca Tiganului och olika delar av Rumänien, medan refugierna för ek, lind, Steregoiu, belägna 730 och 800 m ö h i Gutaiuluibergen i hassel och al ännu inte är kända. Fördelning och föränd- nordvästra Karpaterna. Analys av pollen, växtmakro- ringar (invandring, expansion, regression) mellan olika fossil, kolpartiklar, organisk kol och mineralmagnetiska trädslag var troligen klimatisk betingade under slutfasen parametrar, samt en detaljerad tidsskala (AMS 14C date- av sista istiden. Skogsutvecklingen under de sista 11 ringar av terrestriska växtmakrofossil) möjliggör en nog- 500 åren har däremot mer styrts av konkurrens mellan grann rekonstruktion av miljöförändringarna under slut- arterna, olika spridningshastighet, jordmånsförhållanden fasen av sista istiden och under den nuvarande mellan- och till viss grad också växlande klimatförhållanden. istiden (Holocen). Omkring 14 700 år före nutid (BP) ersattes den tidi- gare öppna busk- och örtvegetationen av gles skog med Rezumat în limba românã dominans av tall och björk, som en följd av en markant klimatförbättring. Under de efterföljande cirka 300 åren Cu toate cã ºcoala palinologicã are o veche tradiþie în (14 100 – 13 800 år BP) blev klimatet kallare och skogen România, neajunsuri ca rezoluþia stratigraficã ºi ecologicã glesnade, medan gräsmark expanderade. När klimatet åter scãzutã, precum ºi lipsa totalã a datelor de cronologie förbättrades spred sig gran och bildade skogar tillsam- absolutã, a fãcut dificilã interpretarea profilelor polenice mans med tall, björk, al och lärk. Mellan 12 600 – 11 500 år în sensul stabilirii compoziþiei vegetale de-a lungul BP förändrades vegetationens sammansättning kraftigt, timpului med en ökning av örter och buskar, och små glesa skogs- Studiul de faþa îºi propune reconstituirea vegetaþiei bestånd av lärk, tall, björk och al. Förändringen skedde ºi a climatului din ultimii 15.000 de ani, ceea ce ar troligen som respons på kalla och torra klimatförhållan- corespunde ultimei pãrþi a glaciarului Weichsel (Würm), den. tranziþia Pleistocen-Holocen ºi mai cu seamã Holocenul, Vid övergången till den nuvarande mellanistiden accentuându-se urmãtoarele aspecte: (Holocen), för cirka 11 500 år sedan expanderade tall, • Evidenþierea existenþei unor refugii glaciare björk och lärk mycket snabbt, som en följd av varmare pentru diferiþi arbori în Romania, în particular , och mer fuktiga klimatförhållanden. Bara 250 år senare în Munþii Gutâiului. bildades täta skogar bestående av tall, björk, lärk, alm • Dinamica ºi factorii care au controlat dinamica och gran. Den snabba spridningen av alm och gran i arborilor de la sfârºitul glaciaþiunii ºi pânã în början av Holocen visar att båda arter troligen fanns prezent. kvar i mindre bestånd under den föregående kallfasen. • Fluctuaþiile climatice reflectate în schimbãrile Almen var det ädla lövträd som spreds snabbast, senare apãrute în comunitatea vegetalã. inom några århundraden följd av ek, lind, ask, lönn och hassel. De artrika lövskogarna, som dominerade under Reconstituirile de paleomediu s-au efectuat pe flancul drygt 2000 år, indikerar troligen ett mer kontinentalt kli- vestic al Munþilor Gutâi, pe baza sedimentelor acumulate mat. Därefter sker en förskjutning till dominans av gran în douã vechi cratere vulcanice, Steregoiu (790 m alt.) ºi och hassel med antydan om en återgång till ett kallare Preluca Þiganului (730 m alt.). Investigaþiile au fãcut uz klimat. De trädslag som invandrar och sprider sig sist de analiza polinicã ca metodã de bazã, fiind pentru prima (för drygt 5000 år sedan) är avenbok och bok. Mänsklig datã completatã cu analize de macro-resturi vegetale ºi påverkan på vegetationen kan spåras först för 2300 år animale, micro-particule de cãrbune, litostratigrafie, ma- sedan, och det är bara under de senaste 300 åren som terial magnetic mineral ºi studii privind cantitatea de car- avverkning och bete varit betydande i den här delen av bon organic. Cronologia a fost stabilitã pe baza Rumänien. numeroaselor datãri 14C AMS. Analizele s-au efectuat la

40 o rezoluþie temporalã ºi spatialã ridicatã, fapt ce a permis Carpinus ºi Fagus apar ºi dezvoltã pãduri mult mai detectarea unor schimbãri de vegetaþie, precum ºi târziu, în jur de 5.700, respectiv 4.800 ani BP, cel mai fluctuaþii climatice de scurtã duratã. probabil ca urmare a schimbãrilor climatice ºi a pãturii de sol. Influenþa antropicã nu este totuºi exclusã, deºi Urmãrind dinamica vegetaþiei de-a lungul informaþiile de care dispunem nu sunt suficient de Tardiglaciarului, putem concluziona cã fluctuaþiile sugestive. Influenþa antropicã este semnalatã acum 2.300 climatice din aceastã perioadã, precum ºi toleranþa de ani, pentru ca în ultimii 300 de ani sã devinã foarte fiziologicã a diverºilor taxoni au avut un rol esenþial. pronunþatã. Acest impact se reflectã în extinderea Astfel, ansamblul polenic ºi cel al macroplantelor prezent pajiºtilor, a pãºunatului în pãdure, existând dovezi privind în sedimentele depuse înainte de 14.700 ani BP indicã o activitãþi de defriºare, ceea ce a contribuit la reducerea vegetaþie lipsitã de arbori, dominatã în schimb de ierburi diversitãþii ºi densitãþii covorului silvestru. ºi arbuºti rezistenþi la condiþii de climat rece ºi arid. Studiul de faþã pune în evidenþã prezenþa în Munþii Creºterea temperaturilor ºi a precipitaþiilor de acum Gutâi de-a lungul Tardiglaciarului, a urmãtoarelor conifere ~14.700 de ani corespunde cu începutul Tardiglaciarului ºi foiose rezistente la climat rece: Pinus, Betula, Larix ºi a determinat apariþia ºi expansiunea gradatã a vegetaþiei Picea, Salix ºi Alnus. Prezenþa polenului de Ulmus între forestiere dominatã de Pinus ºi Betula, ºi la stabilizarea 13,200 ºi 12,900 de ani BP, coroboratã cu expansiunea sa versanþilor. Între 14.100–13.800 ani BP vegetaþia dominatã rapidã înregistratã la începutul Holocenului, poate indica de arbuºti ºi ierburi s-a extins din nou, în timp ce covorul faptul cã ulmul a fost prezent în intervalul 13,400–12,900 forestier s-a redus la exemplare rãzleþe Pinus, Betula ºi ani BP ºi a supravieþuit epsisodului rece (desfaºurat între Larix, indicând reinstalarea unor condiþii climatice mai 12,900–11,500) ani BP în apropierea locului de studiu. În reci. Un nou episod cald desfãºurat între 13.800–12.900 schimb, prezenþa izolatã a granulelor polenice de Quercus, ani BP este evidenþiat de expansiunea semnificativã a Tilia, Fraxinus ºi Corylus este insuficientã pentru a pãdurilor dominate de Picea, la care se adaugã Betula, dovedi supravieþuirea acestor arbori în apropiere Larix, Pinus ºi Alnus. Ulmus, probabil forma populaþii arealului investigat. mici pe versanþii din împrejurimile siturilor studiate. Perioada cuprinsã între 12.950–11.500 ani BP este caracterizatã de retragerea taxonilor forestieri - începând References cu Ulmus, urmat apoi de Picea, Betula ºi Pinus -, în paralel cu expansiunea rapidã a ierburilor ºi arbuºtilor, Ammann, B. 1989. Late-Quaternary palynology at constituind un indiciu al instalãri unui nou episod rece ºi Lobsigensee. Regional vegetation history and local lake devel- mai puþin umed. opment. Dissertationes Botanicae. J. Cramer Verlag Berlin- Tranziþia Tardiglaciar/Holocen are loc acum 11.500 Stuttgart 157 p. ani BP, iar evoluþia vegetaþiei postglaciare a fost un proces Ammann B, Birks HJB, Brooks SJ, Eicher U, von Grafenstein complex, rezultat al interferenþei unor factorii precum, U, Hofmann W, Lemdahl G, Schwander J, Tobolski K, Wick L. clima, procesele de succesiune vegetalã, migraþia 2000. Quantification of biotic response to rapid climatic changes arborilor, competiþia interspecificã, solurile sau impactul around the Younger Dryas-a synthesis. Palaeogeography, antropic. Astfel, expansiunea urmãtorilor taxoni lemnoºi: Palaeoclimatology, Palaeoecology 159: 313-347. Betula, Larix, Pinus ºi Alnus de acum 11.500 ani BP poate Andersen ST. 1974. The Eemian freshwater deposits at fi interpretatã ca fiind un proces succesional, întrucât Egernsund, South Jylland, and the Eemian ladnscape develop- aceºti arbori au fost prezenþi în împrejurimi de-a lungul ment in Denmark. Danmarks Geologiske Undersøgelse Årbøg: Tardiglaciarului. În timp foarte scurt (250 ani BP) a urmat 29-70. expansiunea taxonilor Ulmus ºi Picea, care, probabil au Arztegui D, Asioli A, Lowe JJ, Trincardi F, Vigliotti L, supravieþuit în habitate favorabile destul de aproape de Tamburini F, Chondrogianni C, Accorsi CA, Bandini Mazzanti bazinele de sedimenate. Apariþia regionalã a foioaselor M, Mercuri AM, van der Kaars S, McKenzie JA, Oldfield F. mezo-termofile Quercus, Tilia, Fraxinus ºi Corylus s-a 2000. Palaeoclimate and formation of sapropel S1: inferences produs probabil acum 11.250 de ani BP, pentru ca from Late Quaternary lacustrine and marine sequences in the expansiunea localã sã înceapã în urmã cu 10.700 de ani central Mediterranean region. Palaeogeography, BP când covorul forestier devine diversificat ºi mult mai Palaeoclimatology, Palaeoecology 158: 215-240. dens. Aceastã întârziere are ca ºi cauzã probabilã migraþia Autio J, Hicks S. 2004. Annual variations in pollen deposition lor din refugii mai îndepãrtate. Cinci sute de ani mai târziu and meteorological conditions on the fell Aakenustunturi in (~10.200 de ani BP) Corylus se stabileºte în Munþii Gutâi, northern Finland: Potential for using fossil pollen as a climate pentru ca în jur de 9.300 ani BP sã devinã taxonul domi- proxy. Grana 23:3 1-47. nant. Competiþia interspecificã dintre taxonii prezenþi deja Baldini JUL, McDermott F, Fairchild IJ. 2002. Structure of în regiune ºi noii sosiþi, la care se adaugã noile condiþii the 8,200 cold event revealed by a speleothem trace element climatice, au jucat un rol important în reorganizarea record. Science 296: 2223-2226. forestierã ºi în formarea etajelor de vegetaþie. Balteanu D, Ielenicz M, Popescu N. 1998. Geomorphology of

41 the Romanian Carpathians. New trends and evolutions. Studia lacustrine responses. Journal of Quaternary Science 12: 455- Geomorphologica Carpatho-Balcanica 32: 89–109. 465. Behre KE. 1966. The Late Glacial and early postglacial history Björck S, Walker MJC, Cwynar LC, Johnsen S, Knudsen KL, of vegetation and climate in north-western Germany. Review Lowe JJ, Wohlfarth B, INTIMATE members. 1998. An event of Palaeobotany and Palynology 4:149-161. stratigraphy for the Last Termination in the North Atlantic Behre KE. 1981. The Interpretation of Anthropogenic Indica- region based on the Greenland ice-core record: a proposal by tors in Pollen Diagrams. Pollen et Spores 213: 225-245. the INTIMATE group. Journal of Quaternary Science 13: Behre KE. 1988. The role of Man in European vegetation 283–292. history. In: Huntley B, Webb T III (eds.). Vegetation History. Björck S, Muscheler R, Kromer B, Andersen CS, Heinemeier Kluwer Academic Publishers: 633-672. J, Johnsen S, Conley D, Koc N, Spurk M. Veski S. 2001. High- Bengtsson L, Enell M. 1986. Chemical analysis. In: Berglund resolution analysis of an early Holocene climate event may BE (ed.). Handbook of Holocene Palaeoecology and imply decreased solar forcing as an important climate trigger. Palaeohydrology.Wiley & Sons, Chichester: 423-454. Geology 12: 1107-1110. Bennett KD, Tzedakis PC, Willis KJ. 1991. Quaternary refugia Björkman L, Feurdean A, Cinthio K, Wohlfarth B, Possnert G. of north European trees. Journal of Biogeography 18: 103- 2002. Lateglacial and early Holocene vegetation development 115. in the Gutaiului Mountains, NW Romania. Quaternary Sci- Bennett KD, Willis KJ. 1995. The role of ecological factors in ence Reviews 21: 1039-1059. controlling vegetation dynamics on long temporal scales. Björkman L, Feurdean A, Wohlfarth B. 2003. Lateglacial and Giornale Botanico Italiano 129, 8: 243-254. Holocene forest dynamics at Steregoiu in the Gutaiului Moun- Bennett KD, Willis KJ. 2001. Pollen. In: Smol JP, Birks HJB, tains, NW Romania. Review of Palaeobotany and Palynology Last WM (eds.). Tracking environmental change using lake 124: 79-111. sediments. Kluwer Academic Publishers: 5-32. Blytt, A. 1881. Die Theorie der wechselnden kontinentalen Berglund BE. 2000. The Ystad Project—a case study for und insularen Klimate. Englisches Botanisches Jahrbuch 2: 1- multidisciplinary research on long-term human impact. PAGES 50. Newsletter 8: 6–7. Bodnariuc A, Bouchette A, Dedoubat JJ, Otto T, Fountugne Berglund BE. 2003. Human impact and climatic changes-syn- M, Jalut G. 2002. Holocene vegetational history of the chronous events and a causal link? Quaternary International Apuseni mountains, central Romania. Quaternary Science 105: 7-12. Reviews 21: 1465-1488. Berglund BE, Ralska-Jasiewiczowa M. 1986. Pollen analysis Bond G, Showers W, Chesesby M, Lotti R, Almasi P, and pollen diagrams. In: Berglund BE. (ed.). Handbook of deMenocal P, Priore P, Cullen H, Hajdas I, Bonani G. 1997. A Holocene Palaeoecology and Palaeohydrology. Wiley & Sons, pervasive millennial-scale cycle in North Atlantic Holocene Chichester: 455–484. and glacial climates. Science 278: 1257-1266. Birks HH. 1973. Modern macrofossil assemblages in lake Bond G, Kromer B, Beer J, Muscheler R, Evans MN, Shower sediments in Minnesota. In: Birks HJB, West RG (eds.). Qua- W, Hoffman S, Lotti-Bond R, Hajdas I, Bonani G. 2001. Per- ternary Plant Ecology. Blackwell Scientific Publications, Ox- sistent solar influence on North Atlantic climate during the ford: 173-189. Holocene. Science 294: 2130-2136. Birks HH. 2001. Plant macrofossils. In: Smol JP, Birks HJB, Borcos M, Peltz S, Stan N, Udrescu C, Vasiliu C. 1979. Last WM (eds.). Tracking environmental change using lake Petrochimical and geochemical remarks on Neogen volcanics sediments. Kluwer Academic Publishers: 49-74. from Oas Mountains. Studii Tehnice si Economice, Insitutul de Birks HH. 2003. The importance of plant macrofossils in the Geologie si Geofizica, Mineralogie-Petrolografie 16: 5-33. reconstruction of Lateglacial vegetation and climate: examples Boscaiu N, Lupsa V. 1967. Cercetarii palinologice in Pestera from Scotland, western Norway, and Minnesota, USA. Qua- lui Veterani din Defileul Dunarii. Contributii Botanice: 39-46. ternary Science Reviews 22: 453-473. Boscaiu N, Lupsa V, Olos E, Panzaru G. 1983. Aspecte din Birks HH, Ammann B. 2000. Two terrestrial records of rapid trecutul vegetatiei muntilor Rodnei. in: Rezervatia naturala climatic changes during the glacial-Holocene transition (14,000- Pietrosul Rodnei la 50 de ani de la infiintare. Lucrarile 9,000 calendar years B.P.) from Europe. Proceedings of the Simpozionului aniversar 2-4 sept. 1982, Cluj-Napoca: 232- National Academy of Science of USA 97, 4: 1390-1394. 249. Birks HH, Birks HJB. 2000. Future use of pollen analysis Bozilova ED. 1996. Type region BG-e, Vitosha, Rila, Pirin and must include plant macrofossils. Journal of Biogeography 27: Rodophes Mountains. In: Berglund B, Birks HJB, Ralska- 31-35. Jasiewiczowa M, & Wright H. (eds.). Palaeoecological events Birks HJB. 1986. Numerical zonation, comparison and corre- during the last 15000 years. John Wiley & Sons, Chichester: lation of Quaternary pollen startigraphical data. In: Berglund 719-721. BE. (ed.). Handbook of Holocene Palaeoecology and Bozilova ED, Tonkov SB. 2000. Pollen from Lake Sedmo Palaeohydrology. Wiley & Sons, Chichester: 455–484. Rilsko reveals southeast European postglacial vegetation in Björck S, Rundgren M, Ingolfsson O, Funder S. 1997. The the highest mountain area of the Balkans. New Phytologist 148: Preboreal oscillation around the Nordic Seas: terrestrial and 315–325.

42 Bronk Ramsey C. 1995. Radiocarbon calibration and analysis glaciares, a la lumiere des analyses palynologiques de sequences of stratigraphy: the OxCal program. Radiocarbon 37: 425– datees 14C. Contributii Botanice: 275-287. 430. Donita N. 1962. Elemente pentru interpretarea zonalitatii Brooks SJ, Mayle FE, Lowe JH. 1997. Chironomid-based Late vegetatiei din R.P.R. Acta Botanica Horti Bucurestiensis: 919- glacial climatic reconstruction for southeast Scotland. Journal 935. of Quaternary Science 12: 161-167. Edelstein O, Istvan D, Cojocea C, Weisz G, Bernad A, Stan Broström A. 2002. Estimating source area of pollen and pollen D, Kovacs M. 1980. Geological map of Oas-Tibles Mts., sc. productivity in the cultural landscape of southern Sweden- 1:25.000. Arhiva S.C. IPEG Maramures S.A. Baia Mare (in developing a palynological tool for quantifying past plant cover. Romanian). PhD thesis LUNDQUA 46, Quaternary Geology, Lund Uni- Erdtman G. 1934. Über die Verwendung von versity 116 p. Essigsäureanhydrid bei Pollenuntersunchungen. Svensk Buz Z. 1986. Analize de polen in turba de la Dealu (jud. Botanisk Tidskrift 28: 354-358. Harghita). Contributii Botanice: 95-99. Faegri K. 1963. Problems of immigration and dispersal of the Buz Z. 1999. Cercetari fitosociologice si palinologice in zona Scandinavian flora. North Atlantic Biota and their History: 221- Sovata-Praid-Dealu. PhD thesis, Babes-Bolyai, Cluj Univer- 232. sity. Casa Cartii de Stiinta 214 p. Faegri K, Iversen J. 1989. Textbook of pollen analysis (4th Ciobanu I. 1948. Analiza de polen in turba masivului Semenic. ed.). Wiley Chichester 329 p. Dept. of Biology, Cluj. Farcas S. 1995. The Palynological Analysis of the Peat Bogs Ciobanu I. 1958. Analiza polenica a turbei de la Muhla (M. Complex in Cica Mica, Caliman Mountains. Studia Universi- Apuseni). Contributii Botanice: 246-255. tatis Babes-Bolyai, ser. Geologie 2: 83-94. Ciobanu I. 1960. Cercetarii polenanalitice in Muntii Retezatului. Farcas S. 1996. Istoria vegetatiei din Carpatii romanesti, Studia Universitatis Babes-Bolyai ser. Biologia 5: 47-63. oglindita in analizele palinologice de la Iezerul Caliman, Muntii Constantin S. 2003. Evolutii paleoclimatice in Cuaternar pe Caliman. Contributii Botanice: 83-92. baza speleotemelor din carstul Muntilor Banatului si Mehedinti Farcas S. 2001. Pollen data on peat bogs in the Caliman Moun- Unpublished PhD thesis, Dept. of Geology, Bukarest Univer- tains. Unpublished PhD thesis, Babes-Bolyai University, De- sity. partment of Geology, Cluj 256 p. Cristea V. 1993. Fitosociologie si vegetatia Romaniei. Babes - Farcas S, de Beaulieu JL, Reille M, Coldea G, Diaconeasa B, Bolyai University, Cluj 314 p. Goeury C, Goslar T, Jull T. 1999. First 14C dating of Late Csurös M. 1976. Vegetatia R.S. Romania, cu elemente de Glacial and Holocene pollen sequences from the Romanian ocotirea naturii. Babes-Bolyai University, Cluj. Carpathians. Comptes Rendues de l’Académie des Sciences de Denèfle M, Lézine AM, Fouache E, Dufaure JJ. 2000. A Paris, Sciences de la Vie 322: 799–807. 12,000-year pollen record from Lake Maliq, Albania. Quater- Farcas S, Tantau I. 2002. L’evolution de la vegetation tardi-et nary Research 54: 423–432. postglaciare de Monts Calimani dans le contexte de celle de Diaconeasea B. 1969. Mlastinile de turba din zona alpine a Roumanie. Contributii Botanice: 265-274. Muntilor Fagaras. Importanta lor fitoistorica. Unpublished PhD Feurdean A. in press. Holocene forest dynamics in north-west- thesis. Bukarest University. ern Romania. The Holocene. Diaconeasa B. 1977. Valoarea documentara fitoistorica a Firbas F. 1949. Spät- und nacheiszeitliche Waldgeschichte mlastinii de turba de la Mangalia-Herghelie (jud. Constanta). Mitteleuropas nördlich der Alpen. Band I. Fischer Verlag, Jena Contributii Botanice: 41-53. 480 p. Diaconeasa B. 1979. Analiza sporo polinica a stratului turbos Firbas F. 1952. Spät- und nacheiszeitliche Waldgeschichte de la Calea Magherus (jud. Bistrita). Contributii Botanice: 77- Mitteleuropas nördlich der Alpen. Band II. Fischer Verlag, 83. Jena 256 p. Diaconeasa B, Guist-Homm E. 1981. Cercetari palinologice Grimm EC. 1987. CONISS: a FORTRAN 77 program for privind padurile postglaciare de pe Valea Morii-Cluj Napoca. stratigraphically constrained cluster analysis by the method of Contributii Botanice: 17-24. incremental sum squares. Computers and Geosciences 13: 13- Diaconeasa B, Buz Z. 1993. Analiza palinologica a mlastinii 37. Podul de Hartie (Corund, jud. Harghita). Studia Universitatis Grimm EC. 1992. Tilia and Tilia-graph: Pollen spreadsheet Babes-Bolyai, ser. Biologia 38: 31-37. and graphics programs. Programs and Abstracts, 8th Interna- Diaconeasa B, Farcas S. 1995-1996. Stejarisurile amestecate, tional Palynological Congress, Aix-en-Provence, September evolutia si dinamica lor in Pleistocenul din Romania. Contributii 6-12, 1992: 56. Botanice: 93-102. Haesaerts P, Borziak I, van der Plicht I, Damblon F. 1998. Diaconeasa B, Farcas S. 1998. Particularitati ale evolutiei Climatic events and upper Paleolithic chronology in the Dniester vegetatiei in sud-estul Romaniei, comparativ cu Transilvania, basin: New 14C results from Cosautsi. Radiocarbon 40: 649- relevate prin analize palinologice. Studia Universitatis Babes- 657. Bolyai, ser. Biologia 1-2: 27-36. Hannon GE. 1999. The use of plant macrofossils and pollen in Diaconeasa B, Farcas S. 2002. Aspects concernant les refuges the palaeoecological reconstruction of vegetation. PhD thesis

43 SILVESTRIA 106, Swedish University of Agricultural Science. 21,000 years. Quaternary Science Reviews 17: 473-506. Hannon GE, Bradshaw RHW, Wastegård S. 2003. Rapid veg- Lang G. 1970. Florengeschichte und mediterranmitteleuropaishe etation change during the early Holocene in the Faeroe Islands Florenbeziehungen. Feddes Rep. 81: 315-335. detected in terrestrial and aquatic ecosystems. Journal of Qua- Lang G. 1994. Quartäre Vegetationsgsechichte Europas; Fischer ternary Science 18: 615-619. Verlag: Jena 462 p. Hicks S. 1994. Present and past pollen records of Lapland Lascu I. 2003. Studiul Tardiglaciarului de la Sieu-Magherus forests. Review of Palaeobotany and Palynology 82: 17-35. (jud. Bistrita-Nasaud). Unpublished master thesis, Babes- Hicks S. 2001. The use of annual arboreal pollen deposition Bolyai University, Department of Geology, Cluj Napoca. values for delimiting tree-lines in the landscape and exploring Lowe JJ. 2000. Abrupt climatic changes in Europe during the models of pollen dispersal. Review of Palaeobotany and last glacial-interglacial transition: the potential for testing hy- Palynology 117: 1-29. potheses on the synchronicity of climatic events using Hoek W. 2001. Vegetation response to the 14.7 and 11.5 ka cal. tephrochronology. Global and Planetary Change 30: 73-84. BP climate transitions: is vegetation lagging climate? Global Lupsa V. 1972. Analiza sporo-polinica a mlastinii de la Zagra and Planetary Change 30: 103-115. (jud. Bistrita-Nasaud). Studii si Cercetari de Biologie, Seria Huntley B, Birks HJB. 1983. An Atlas of Past and Present Botanica 24: 363-366. Pollen Maps for Europe: 0–13000 years ago; Cambridge Uni- Lupsa V. 1977. Cercetari palinologice in mlastinile de turba din versity Press: Cambridge. rezervatia Cornedei-Ciungii Balasinii (Muntii Maramuresului). Istvan D, Popescu S, Pop I. 1990. Muntii Gutai. Editura Sport Ocrotirea Naturii Suceava 4: 291-301. Turism: Bucuresti 149 p. Lupsa V. 1980. Evolutia postglaciara a ecosistemelor forestiere Iversen, J. 1941. Landnam i Danmarks Stenalder. Danmarks din “Tara Oasului” (Jud. Satu Mare). Contributii Botanice: 63– Geologiske Undersøgelse Række II 66: 1-68. 68. Iversen J. 1942. En pollenanalytisk Tidsfaestelse af Magny M, Begeot C, Guiot J, Peyron O. 2003. Constraining Ferkvandslagene ved Nørre Lyngby. Dansk Geologisk Fören- patterns of hydrological changes in Europe in response to ing Medd. 10, Copenhagen . Holocene climate cooling phases. Quaternary Science Reviews Iversen, J. 1949. The influence of prehistoric man on vegeta- 22: 1589-1596. tion. Danmarks Geolgiske Undersøgelse Række IV 3 (6): 1- Marian M. 1999. Vegetatia Oasului. In: Flora si fauna Oasului. 25. Ardelean G, Marian M. (eds.). Gutenberg Press: Arad . Iversen J. 1954. The Late-glacial flora of Denmark and its Moore P., Webb JA, Collinson ME. 1991. Pollen analysis, 2nd relation to climate and soil. Danmarks Geologiske edn. Blackwell Scientific Publications, Oxford. Undersøgelse Årbøg 2: 87-119. Nesje A, Dahl OS. 2001. The Greenland 8200 cal. yr BP event Iversen, J. 1960. Problems of the early postglacial forest in detected in loss-on-ignition profiles in Norwegian lacustrine Denmark. Danmarks Geologiske Undersøgelse 4/3: 1-32. sediment sequences. Journal of Quaternary Science 16: 155- Jacobson GI, Bradshaw RJW. 1981. The selection of sites for 166. paleovegetational studies. Quaternary Research 16: 80-96. Nesje A, Matthews JA. Dahl SO, Berrisford MS, Andersson Karácsonyi C. 1995. Flora si vegetatia judetului Satu Mare: C. 2001: Holocene glacier fluctuations of Flatebreen and win- Editura Muzeului Satmarean 181 p. ter-precipitation changes in the Jostedalsbreen region, western Kasse C, Huijzer AS, Krzyszkowski D, Bohncke SJP, Coope Norway, based on glaciolacustrine sediment records. The GR. 1998. Weichselian Late Pleniglacial and Late-glacial Holocene 11: 267-280. depositional environments, Coleoptera and pleniglacial climatic Nilsson T. 1935. Die pollenanalytische Zonengliederung der records from central Poland (Belchowa). Journal of Quater- spät-und postglazialen Bildungen Schonens. Geologiska nary Science 13: 455-469. Föreningens i Stockholm Förhandlingar 57: 385-562. Katz NJ, Katz SV, Kipiani MG. 1965. Atlas and keys of fruits Nilsson T. 1961. Ein neues Standardpollendiagramm aus and seeds occurring in the Quaternary deposits of the U.S.S.R. Bjärsjöholmssjön in Schonen. The Institutes of Mineralogy, Nauka, Moscow 356 p. Paleontology and Quaternary Geology, University of Lund, Kotarba A, Baumgart-Kotarba M. 1999. Problems of glacia- Sweden 96 p. tions of the High Tatra Mountains. Zeitschrift für Onac BP, Lauritzen SE. 1996. The climate of the last 150.000 Geomorphologie NF 113: 19-31. years recorded in speleothems: preliminary results from north- Kullman L. 1998. The occurrence of thermophilous trees in western Romania. Theoretical and Applied Karstology 9: 9-21. the Scandes Mountains during the early Holocene: evidence Onac BP, Constantin S, Lundberg J, Lauritzen SE. 2002. Iso- for a diverse tree flora from macroscopic remains. Journal of topic climate record in a Holocene stalagmite from Ursilor Ecology 86: 421-428. Cave (Romania). Journal of Quaternary Science 17: 319-327. Kullman L. 2002. Boreal tree taxa in the central Scandes during Peteet D. 2000. Sensitivity and rapidity of vegetational re- the Late-Glacial: implications for Late Quaternary forest his- sponse to abrupt climate change: Proceedings of the National tory. Journal of Biogeography 29: 1117-1124. Academy of Science, USA 97: 1359-1361. Kutzbach JE, Galimore R, Harrison SP, Behling P, Selin R, Pop E. 1929. Analize de polen in turba Carpatilor Orientali Laarrif F. 1998. Climate and biome simulation for the past (Dorna-Lucina). Buletinul Gradinii Botanice si al Muzeului

44 Botanic, Cluj 9: 81-210. 177. Pop E. 1932. Contributii la istoria vegetatiei cuaternare din Reid C. 1899. The origin of the British Flora. Dulau & Co., Transilvania. Buletinul Gradinii Botanice si al Muzeului Botanic, London. Cluj 12: 29-102. Reille, M. 1992. Pollen et spores d’Europe et d’Afrique du Pop E. 1942. Contributii la istoria padurilor din nordul nord. Laboratoire de botanique historique et palynologie, Transilvaniei. Buletinul Gradinii Botanice si al Muzeului Bo- Marseilles 520 p. tanic, Cluj 22: 101-177. Reille M. 1995. Pollen et spores d’Europe et d’Afrique du Pop E. 1957. Analize de polen in regiuni de campie. Buletin nord. Supplement 1. Laboratoire de botanique historique et stiintific, Sectia de Biologie si Stiinte Agricole, Seria Botanica palynologie, Marseilles 329 p.’ 9: 5-32. Renssen H, Isarin RFB. 2001. The two major warming phases Pop E. 1958. Regiunea de mlastini eutrofe Dragoiasa-Bilbor- of the last deglaciation at 14.7 and 11.5 ka cal BP in Europe: Borsec si importanta ei fitogeografica. Ocrotirea Naturi 3: 11- climate reconstructions and AGCM experiments. Global and 42. Planetary Changes 30: 117-153. Pop E. 1960. Mlastinile de turba din R.P.R. Editura Academiei Renssen H, Isarin RFB, Vandenberghe J. 2001. Rapid climatic Romane, Bucuresti 515 p. warming at the end of the last glacial new perspectives. Global Pop E. 1971. Primele determinari de varsta cu radiocarbon in and Planetary Change 30: 155-165. turba noastra cuaternara. Progrese in palinologia romaneasca Rösch M, Fischer E. 2000. A radiocarbon dated Holocene pol- Editura Academiei Republicii Socialiste Romania, Bukarest: len profile from the Banat Mountains (southwestern 193-204. Carpathians, Romania). Flora 195: 277–286. Pop E, Diaconeasa B. 1964. Analiza palinologica a unui profil Rubnßèkova and Rybnßèek. 1996. In: Berglund BE, Birks HJB, de turba glaciara din raionul Zalau. Studii si Cercetari de Biologie Ralska-Jasiewiczowa M, Wright HE (eds.). Palaeoecological ser. Botanica 16: 259-262. events during the last 15000 years: Regional syntheses of Pop E, Diaconeasa B, Boscaiu N. 1965a. Analiza polinica a palaeoecological studies of lakes and mires in Europe. Wiley turbei de la Taul Negru (rn. Lapus). Studia Universitatis Babes- & Sons, Chichester: 403–472. Bolyai, ser. Biologia 1: 37-39. Rudner ZE, Sümegi P. 2001. Recurring taiga forest-steppe Pop E, Diaconeasa B, Boscaiu N. 1965b. Analiza polinica a habitats in the Carpathian Basin in the Upper Weichselian. turbei din tinovul Tau la Gutai. Natura, ser. Biologia 17: 74- Quaternary International 76/77: 177–189. 77. Seppä H, Birks HH, Birks HJB. 2002. Rapid climatic changes Pop E, Boscaiu N, Diaconeasa B, 1965c. Analiza polinica a during the Greenland stadial 1 (Younger Dryas) to early turbei de la Tau-Baitii (raionul Visau, reg. Maramures). Studia Holocene transition on the Norwegian Barents Sea coast. Boreas Universitatis Babes-Bolyai, ser. Biologia 2: 37-40. 31: 215-225. Pop E, Diaconeasa B. 1967. Analiza palinologica a turbei din Sernander R. 1890. Om förekomsten af subfossila stubbar på tinovul Mohos (Tusnad), Contributii Botanice: 297-303. svenska insjöars botten. Botaniska Notiser, Lund: 10-20. Pop E, Ciobanu I. 1957. Analize de polen in turba de la Cotul Stefanova I, Ammann B. 2003. Lateglacial and Holocene veg- Carpatilor, Buletinul Universitati V. Babes si Bolyai Cluj, Ser. etation belts in the Pirin Mountains (southwestern Bulgaria). Stiinte Naturale 1 (1-2): 453-471. The Holocene 13: 97-107. Prentice IC. 1988. Records of vegetation in time and space: the Stockmarr J.1991. Tablets with spores used in absolute pollen principles of pollen analysis. In: Huntley B, Webb T (eds.). analysis. Pollen et Spores 13: 615-621. Vegetation history, Handbook of vegetation Science 7. Kluwer Sugita S. 1994. Pollen representation of vegetation in Quater- Academic Publishers: 17-42. nary sediments: theory and method in patchy vegetation. Jour- Prentice IC, Sykes MT, Cramer W. 1993. A simulation model nal of Ecology 82: 881-897. for transient effect of climate change on forest landscapes. Sugita S, Gaillard MJ, Brostöm A. 1999. Landscape openess Ecological. Modelling 65: 51-70. and pollen records: a simulation approach. The Holocene 9: Punt W, Clarke CS. 1984. The northwest European pollen 409-421. flora IV. Elsevier Scientific Amsterdam 138 p. Sümegi P, Rudner ZE. 2001. In situ fragments as remains of Ralska-Jasiewiczowa M, Latalowa M. 1996. Poland. In: natural wild fires in the upper Würm of the Carpathians Basin. Berglund BE, Birks HJB, Ralska-Jasiewiczowa M, Wright HE Quaternary International 76/77: 165–176. (eds.). Palaeoecological events during the last 15000 years: Sykes MT, Prentice IC, Cramer W. 1996. A bioclimatic model Regional syntheses of palaeoecological studies of lakes and for the potential distribution of north European tree species mires in Europe. Wiley & Sons, Chichester: 403–472. under present and future climates. Journal of Biogeography Ralska-Jasiewiczowa M, Goslar T, Madeyska T, Starkel L. 23: 203-223. 1998. Lake Gosciaz, central Poland a monographic study. W. Tamas T. 2003. Mineralogy and geochemistry of speleothems Szafer Institute of Botany Polish Academy of Sciences, from some caves in Bihor Mountains. Abstract of the PhD Drukarnia Kolejowa, Krakow 340 p. thesis, Babes-Bolyai University, Department of Geology, Cluj; Ravazzi C. 2001. Late Quaternary history of spruce in south- 36 p. ern Europe. Review of Palaeobotany and Palynology 120: 131- Tamas T, Causse C. 2000. U-Th TIMS chronology of two

45 stalagmites from V11 Cave (Bihor Mountains, Romania). Theo- division, correlation and dating. Global and Planetary Change retical and Applied Karstology 13: 25-32. 30: 59-72. Tantau. 2003. Recherches pollenanalytiques dans les Carpates Walker MJC, Björck S, Lowe JJ, Cwynar LC, Johnsen S, orientales (Roumanie). Historie de la vegetation et de l’action Knudsen KL, Wohlfarth B, and INTIMATE group. 1999. Iso- humaine. Abstract of the PhD thesis, Babes-Bolyai Univer- topic ‘events’ in the GRIP ice core: a stratotype for the Late sity, Department of Geology, Cluj; 29 p. Pleistocene. Quaternary Science Reviews 18: 1143–1150. Tantau I, Reille M, de Beaulieu JL, Farcas S, Goslar T, Paterne Watts WA. 1959. Interglacial deposits at Kilberg and Newtown M. 2003. Vegetation history in the eastern Romanian Co. Waterford. Proceedings Royal Irish Academy 60: 79-134. Carpathians: pollen analysis of two sequences from the Mohos Watts WA. 1978. Plant macrofossils and Quaternary crater. Vegetation History and Archaeobotany 12: 113-125. paleoecology. In: Biology and Quaternary Environments, Tinner W, Lotter A. 2001. Central European vegetation re- Walker D, Guppy JC (eds.). Australian Academy of Science, sponse to abrupt climate change at 8,2 ka. Geology 29: 551- Canberra: 53-67. 554. West RG. 1977. Pleistocene geology and biology and especial Tobolski K, Ammann B. 2000. Macrofossils as records of referenced to the British Isles, 2nd edition. Longman, London plant response to rapid Late Glacial climatic changes at three Wick L. 1996. Late glacial and Early Holocene sites in Swiss Alps. Palaeogeography, Palaeoclimatology, palaeoenvironment in Brianza, N Italy. Il Quarternaria 9: 653- Palaeoecology 159: 251-259. 660. Tonkov S, Panovska H, Possnert G, Bozilova E. 2002. The Wick L. 2000. Vegetational response to climatic changes re- Holocene vegetation history of Northern Pirin Mountains, corded in Swiss Late Glacial lake sediments. Palaeogeography, southwestern Bulgaria: pollen analysis and radiocarbon dating Palaeoclimatology, Palaeoecology 159: 231-250. of a core from Lake Ribno Banderishko. The Holocene 12: Wick L, Lemcke G, Sturm M. 2003. Evidence of Lateglacial 201-210. and Holocene climatic change and human impact in charcoal, Vandenberghe J, Isarin RFB, and Renssen H. 2001. Rapid cli- isotopic and geochemical records from laminated sediments of matic warming: palaeo-data analysis and modeling. Global and lake Van, Turkey. The Holocene 13: 665-675. Planetary Change 30: 1-5. Williams JW, Post DM, Cwynar LC, Lotter AF, Levesque AJ. van der Hammen, Wijmistra T, Zagwijn WH. 1971. The floral 2002. Rapid and widespread vegetation responses to past cli- record of the Late Cenozoic of Europe. In: The late Cenozoic mate change in the North Atlantic region. Geology 11: 971- glacial ages. Turekian KK (ed). Yale University press, new 974. Haven: 391-424. Willis KJ. 1992a. The late Quaternary vegetational history of von Grafenstein U, Erlenkeuser H, Muler J, Jouzel J, Johnsen northwest Greece. I. Lake Gamousti. New Phytologist, 121: S. 1998. The cold event 8200 year ago documented in oxygen 101-117. isotope records of precipitation in Europe and Greenland. Willis KJ. 1992b. The late Quaternary vegetational history of Climate Dynamics 14: 73-81. northwestern Greece. II Rezina marsh. New Phytologist, 121: von Grafenstein U, Erlenkeuser H, Muler J, Brauer A, Jouzel 119-138. J, Johnsen S. 1999. A mid-European decadal isotope-climate Willis KJ. 1994. The vegetational history of the Balkans. Qua- record from 15,500 to 5000 years B.P. Science 284: 1654- ternary Science Reviews 13. 769–788. 1657. Willis KJ. 1996. Where did all flowers go? The fate of temper- von Post L. 1916. Om Skogsträdpollen i sydsvenska ate European flora during glacial period. Endevour 20: 110- torfmosselagerföljder. Geologiska Föreningens i Stockholm 114. Förhandlingar 38: 384-390. Willis KJ, Sümegi P, Braun M, Tóth A. 1995. The late Quater- von Post L. 1946. The prospect for pollen analysis in the nary environmental history of Bátorliget, N.E. Hungary. study of the earth’s climate history. New Phytologist 45: 193- Palaeogeography, Palaeoclimatology, Palaeoecology 118: 25– 217. 47.

Wagner F, Aaby B, Visscher H. 2002. Rapid atmospheric CO2 Willis,KJ, Braun M, Sümegi P, Tóth A. 1997. Does soil change changes associated with the 8,200-years-B.P. cooling event. cause vegetation change or vice versa? A temporal perspective Proceeding of the National Academy of Science of USA, 99, 19. from Hungary. Ecology 78: 740-750. 1211-1214. Willis KJ, Rudner E, Sümegi P. 2000. The full-glacial forests of Wainman H, Mathewes RW. 1990. Distribution of plant macro- central and southeastern Europe: Evidence from Hungarian remains in the surface sediments of Marion Lake, south-west- macrofossil charcoal, pollen and molluscan records. Quaternary ern British Columbia. Canadian Journal of Botany 68: 364- Research 53: 203–213. 373. Wohlfarth B, Hannon G, Feurdean A, Ghergari L, Onac BP. Walden J, Oldfield F, Smith JP. 1999. Environmental Magnet- 2001. Reconstruction of climatic and environmental changes in ism: A Practical Guide. Quaternary Research Association Lon- NW Romania during the early part of the last deglaciation don 250 p. (15,000–13,600 cal years BP). Quaternary Science Reviews Walker MJC. 2001. Rapid climatic change during the last gla- 20: 1897-1914. cial-interglacial transition; implications for stratigraphic sub- Woldstedt P. 1958. Das Eiszeitalter 2. Ferdinand Enke Verlag,

46 Stuttgart, 438 p. Woodward FI. 1988. Temperature and the distribution of plant species. Plants and temperature Long SP, Woodward FI (ed.) Sympo. Soc. Exp. Biol. 42: 59-75. Wright HE Jr. 1984. Sensitivity and response time of natural systems to climatic change in the late quaternary. Quaternary Science Reviews 3: 91-131.

47