holocene vegetation and climate changes in the central

Clim Past 9 2023ndash2042 2013wwwclim-pastnet920232013doi105194cp-9-2023-2013copy Author(s) 2013 CC Attribution 30 License

EGU Journal Logos (RGB)

Advances in Geosciences

Open A

ccess

Natural Hazards and Earth System

Sciences

Open A

ccess

Annales Geophysicae

Open A

ccess

Nonlinear Processes in Geophysics

Open A

ccess

Atmospheric Chemistry

and Physics

Open A

ccess

Atmospheric Chemistry

and PhysicsO

pen Access

Discussions

Atmospheric Measurement

Techniques

Open A

ccess

Atmospheric Measurement

Techniques

Open A

ccess

Discussions

Biogeosciences

Open A

ccess

Open A

ccess

BiogeosciencesDiscussions

Climate of the Past

Open A

ccess

Open A

ccess

Climate of the Past

Discussions

Earth System Dynamics

Open A

ccess

Open A

ccess

Earth System Dynamics

Discussions

GeoscientificInstrumentation

Methods andData Systems

Open A

ccess

GeoscientificInstrumentation

Methods andData Systems

Open A

ccess

Discussions

GeoscientificModel Development

Open A

ccess

Open A

ccess

GeoscientificModel Development

Discussions

Hydrology and Earth System

Sciences

Open A

ccess

Hydrology and Earth System

Sciences

Open A

ccess

Discussions

Ocean Science

Open A

ccess

Open A

ccess

Ocean ScienceDiscussions

Solid Earth

Open A

ccess

Open A

ccess

Solid EarthDiscussions

The Cryosphere

Open A

ccess

Open A

ccess

The CryosphereDiscussions

Natural Hazards and Earth System

Sciences

Open A

ccess

Discussions

Holocene vegetation and climate changes in the centralMediterranean inferred from a high-resolution marine pollenrecord (Adriatic Sea)

N Combourieu-Nebout1 O Peyron2 V Bout-Roumazeilles3 S Goring4 I Dormoy5 S Joannin56 L Sadori7G Siani8 and M Magny5

1CNRSUVSQCEA-UMR 8212 LSCE Gif sur Yvette France2CNRS-UMR 5059 CNRS CBAE-centre de Bio-Archeologie et drsquoEcologie Montpellier France3CNRS-UMR 8217 Geosystemes Universite Lille 1 France4Department of Geography Univ of WisconsinndashMadison Wisconsin 53706 USA5CNRS-UMR 6249 CNRS Laboratoire Chrono-Environnement Besancon France6School of Environment and Development University of Manchester Oxford Road Manchester M13 9PL UK7Dipartimento di Biologia Ambientale Universita La Sapienza Roma Italy8CNRS-UMR 8148 IDES Departement des Sciences de la Terre Universite Paris Sud Orsay France

Correspondence toN Combourieu-Nebout (nathalieneboutlsceipslfr)

Received 26 March 2013 ndash Published in Clim Past Discuss 5 April 2013Revised 10 July 2013 ndash Accepted 12 July 2013 ndash Published 2 September 2013

Abstract The high-resolution multiproxy study of the Adri-atic marine core MD 90-917 provides new insights to re-construct vegetation and regional climate changes over thesouthcentral Mediterranean during the Younger Dryas (YD)and Holocene Pollen records show the rapid forest colo-nization of the Italian and Balkan borderlands and the grad-ual installation of the Mediterranean association during theHolocene Quantitative estimates based on pollen data pro-vide Holocene precipitations and temperatures in the Adri-atic Sea using a multi-method approach Clay mineral ratiosfrom the same core reflect the relative contributions of river-ine (illite and smectite) and eolian (kaolinite) contributionsto the site and thus act as an additional proxy with which toevaluate precipitation changes in the Holocene

Vegetation climate reconstructions show the responseto the Preboreal oscillation (PBO) most likely driven bychanges in temperature and seasonal precipitation which islinked to increasing river inputs from Adriatic rivers recordedby increase in clay mineral contribution to marine sedimentsPollen-inferred temperature declines during the earlyndashmidHolocene then increases during the midndashlate Holocene sim-ilar to southwestern Mediterranean climatic patterns duringthe Holocene Several short vegetation and climatic events

appear in the record indicating the sensitivity of vegeta-tion in the region to millennial-scale variability Recon-structed summer precipitation shows a regional maximum(170ndash200 mm) between 8000 and 7000 similar to the generalpattern across southern Europe Two important shifts in vege-tation occur at 7700 cal yr BP (calendar years before present)and between 7500 and 7000 cal yr BP and are correlated withincreased river inputs around the Adriatic Basin respectivelyfrom the northern (7700 event) and from the central Adriaticborderlands (7500ndash7000 event) During the mid-Holocenethe wet summers lead to permanent moisture all year re-sulting in a homogeneous seasonal precipitation regime Af-ter 6000 cal yr BP summer precipitation decreases towardspresent-day values while winter precipitation rises regularlyshowing the setting up of Mediterranean climate conditions

Multiproxy evidence from core MD 90-917 provides adeeper understanding of the role of precipitation and particu-larly the seasonality of precipitation in mediating vegetationchange in the central Mediterranean during the Holocene

Published by Copernicus Publications on behalf of the European Geosciences Union

2024 N Combourieu-Nebout et al Holocene vegetation and climate changes

1 Introduction

The Mediterranean is highly sensitive to climate change andMediterranean ecosystems already widely affected by an-thropogenic pressures are likely to be strongly impacted byfuture warming and drought stresses Precipitation changesare of particular concern for the region Increasing dry-ness predicted by models for the Mediterranean will in-duce large water resource deficits and might threaten the re-gionrsquos habitability (Giorgi 2006 IPCC 2007 Giorgi andLionello 2008) While modern Mediterranean precipitationis controlled by the influences of both subtropical and mid-latitudinal climatic belts resulting in dry summers and rainywinters climate in the Mediterranean has changed during theHolocene providing an opportunity to study ecosystem re-sponse under a range of climate conditions Thus reconstruc-tions of past vegetation and climate in the different basinsof the Mediterranean will produce information allowing usto draw a general framework to understand regional changeand the drivers of environmental and climate change acrossthe whole basin

The southern Adriatic with its central location in theMediterranean Basin provides an excellent opportunity torecord high-resolution vegetation changes in Italy and theBalkans in response to Holocene climate variability In factcentral Mediterranean ecosystems are highly sensitive to cli-mate changes and help to understand the climate connectionsand conflicting influences between northwestern and south-eastern atmospheric systems and their relative roles in medi-ating precipitation and runoff during the Holocene

Holocene vegetation change in the central Mediterraneanhas been widely investigated using continental and marinerecords (eg Watts et al 1996 Combourieu-Nebout et al1998 Jahns and van den Bogaard 1998 Magri 1999 Ma-gri and Sadori 1999 Allen et al 2002 Oldfield et al 2003Drescher-Schneider et al 2007 Piva et al 2008 Kotthoff etal 2008a b 2011 Di Donato et al 2008 Di Rita and Ma-gri 2009 Joannin et al 2012 2013 Di Rita et al 2013) Al-though these records vary in temporal and taxonomic resolu-tion they show a general pattern of forest expansion linked towarming climates in the Holocene as well as the influence ofmillennial-scale climate variability on continental paleoenvi-ronments Most records from the central Mediterranean di-vide the Holocene into three phases similar to those in otherregions of the Mediterranean (eg Jalut et al 2000 2009Finne et al 2011) with boundaries that have been recentlyrefined by Walker et al (2012) (i) an early Holocene humidperiod from 11 700 to 8200 cal yr BP (calendar years beforepresent) (ii) a mid-Holocene transitional period from 8200to 4200 cal yr BP and (iii) a late-Holocene period of arid-ification following 4200 cal yr BP Climate reconstructionsfrom Lake Accesa (northcentral Italy) distinguish two dis-tinct intervals based on precipitation and emphasize the roleof the seasonality during the Holocene (Peyron et al 2011)Given the scarcity of reliable high-resolution pollen records

in the central Mediterranean and the peculiarities of this re-gion new evidence is needed to better understand the climatechanges and the climate processes in this area

The new pollen record from the marine MD 90-917 coreprovides new insights into vegetation and climate changesfor the central Mediterranean This study part of the Frenchmulti-proxy project LAMA provides a new understanding ofthe links between hydrological variability climate and vege-tation in the central Mediterranean by bringing together con-tinental and marine data to produce an integrated interpre-tation of the paleoenvironmental interpretation of Holoceneclimate changes This paper investigates the impacts oflong-term and millennial-scale climate variability during theHolocene through temperature and precipitation stresses onItalian and Balkan vegetation By examining the clay fractionratios from the same core we will be able to discuss patternsof local and regional runoff linked to precipitation and sea-sonality changes

2 Lithology and age model

Core MD 90-917 (41 N 1737prime E) was collected in the Adri-atic Sea by the RVMarion Dufresneat a depth of 1010 m(Fig 1) The core is described more fully in Combourieu-Nebout et al (1998) and Siani et al (2004 2010) Sedimentsare composed of 21 m of clay interrupted by two black lay-ers corresponding to the deposition of the S1 sapropel in twolevels S1a and S1b respectively at 249ndash255 cm and 229ndash239 cm downcore (Siani et al 2004) The pollen record pre-sented in this paper focuses on the upper 33 m of the core

The age model of the core was previously given in Sianiet al (2010) and uses tephra layers and 21 AMS (accelera-tor mass spectrometry)14C dates performed on monospecificplanktonic foraminifera The conventional radiocarbon ageshave been subsequently converted into calendar ages basedon INTCAL04 (Reimer et al 2004) using the14C calibrationsoftware CALIB 6 The calibration integrates a marine14Creservoir age correctionR(t) at 390plusmn85 yr according to Sianiet al (2000 2001) For this paper only 13 AMS14C are usedand provide a reliable age model for the last 13 000 cal yr BP(Table 1)

3 Present-day environmental settings

31 Climate and atmospheric circulation pattern

Atmospheric circulation in the Adriatic is controlled by twodominant climatic systems In the winter the Eurasian Highextends to the southwest The Eurasian High retreats north-ward in the summer and is replaced by the northeastwardexpansion of the Azores High This pattern results in a typ-ically Mediterranean climate (Fig 1) winters are generallystormy and coolcold while summers are warmhot and drySpring climate in the region is generally long characterized

Clim Past 9 2023ndash2042 2013 wwwclim-pastnet920232013

N Combourieu-Nebout et al Holocene vegetation and climate changes 2025

0

10

20

30

40

50

0

40

80

120

160

200

J F M A M J J A S O N D

BoraMaestral

Jugo

OstroSirocco

Libeccio

ROMA

TUNIS

MD 04-2797

VENICE

4

Trifoglietti

Ledro

Pergusa

Accesa

Maliq

Monticchio

sea surface circulation

marine site continental site

MD 90-917

5

1

62

3

0

10

20

30

40

50

0

40

80

120

160

200


BRINDISI 165 574 10m

BARI 159 586 49m

0

10

20

30

40

50

0

40

80

120

160

200


Monte San Angelo128 613 844m

0

10

20

30

40

50

0

40

80

120

160

200


SPLIT 158 893 128m

0

10

20

30

40

50

0

40

80

120

160

200


DUBROVNIK162 1298 165m

0

10

20

30

40

50

0

40

80

120

160

200


DURRES 159 962 0m

precipitationtemperature

1

Preola

4

2 5

3 6

ombrothermic diagram site

Otranto strait

Po

Fig 1 Location of core MD 90-917 On the map is indicated sea surface circulation in Adriatic Sea and winds blowing over the AdriaticBasin A selection of ombrothermic diagrams is presented on the right for the surrounding borderlands to show the climate of the studied area(Walter et al 1975 New et al 2000 New LoCCllim software) For each ombrothermic diagram near or below the site name is noted theannual mean temperature the total annual precipitation and the altitude of the site For each ombrothermic diagram the left axis correspondsto precipitation values and the right axis to the temperature value Red arrows represent the southndashnorth sea surface water inflow blue arrowsdraw the northndashsouth sea surface outflow

by fluctuating weather while autumn is very short with anabrupt transition toward winter The length and intensity ofsummer dryness increases southwards across the Adriatictotal annual precipitation grades from a maximum in the eastto a minimum in the west (Fig 1) The mountainous regionsin the Adriatic Basin may show patterns that are dramati-cally different than their surroundings In the Apennines orDinarides temperatures decrease and precipitation increasesaccording to altitudinal elevation

A number of winds blow over the central Mediterranean(Fig 1) The bora is a cold wind that commonly blows fromthe north and northeast during winter The mistral (Italianmaestral) is a northwestern wind that blows from spring toautumn more often in summer with decreasing strength tothe south The sirocco blows from the south from autumnto spring bringing dust from North Africa The ostro is asoutherly wind often identified with the sirocco The jugo isan east to southeast dry wind that blows most often in thewinter The west to southwest wind called libeccio blowsyear-round but rarely reaches the Adriatic Basin

32 Hydrology of Adriatic Basin

The hydrology of the Adriatic Sea is seasonally controlled bywinds and river inputs Sea surface circulation follows a gen-eral pattern with a southndashnorth inflow from the Ionian Seathrough the Otranto Basin Flow progresses counterclock-wise from the east coast of the Adriatic to the northndashsouthoutflow along the west coast of the basin The outflow isstrengthened by the discharge from the Po River as well asfrom small rivers running into the Adriatic along the east-ern Italian coast (Fig 1) The eastern Balkans do not con-tribute significantly to fresh water inputs into the Adriaticsince rivers in the region have short runs with low flow pro-ducing only local effects near the coast Thus contributionsto the fresh water and nutrient inputs in the Adriatic Basinare largely derived from Italian river systems dominated byPo River discharges The Po is primarily fed by snowmeltin the spring and high runoff in the autumn particularly inthe northern reaches of the watershed (Cattaneo et al 2003

wwwclim-pastnet920232013 Clim Past 9 2023ndash2042 2013


Table 1Age model for core MD 90-917 AMS14C ages and corresponding calendar ages from INTCAL04 (Reimer et al 2004) integrating14C reservoir age correctionR(t) at 390plusmn 85 yr BP (Siani et al 2000 2001)

Depth (cm) Species 14C age BP (yr) Errorplusmn1σ Cal age BP (yr)

0ndash2 G bulloides 1010 60 555ndash609140ndash142 G ruber 4180 70 4082ndash4290167ndash169 G ruber 4750 70 4855ndash4986175ndash177 G bulloides 5000 70 5344ndash5466190ndash192 G bulloides 5680 70 5990ndash6128230ndash232 G bulloides 6920 90 7413ndash7511240ndash242 G bulloides 7930 80 8171ndash8340250ndash252 G ruber 8170 70 8390ndash8482275ndash277 G bulloides 10 390 90 11 304ndash11 624295ndash297 G bulloides 10 800 90 12 116ndash12 399305ndash307 G bulloides 10 830 90 12 225ndash12 406315ndash317 G bulloides 11 140 90 12 721ndash12 853335ndash337 G bulloides 11 520 100 12 939ndash13 114

Frignani et al 2005 Orange et al 2005 Palinkas and Nit-trouer 2007)

33 Present-day vegetation

Vegetation is organized along altitudinal belts on the Adri-atic borderlands according to the ecological requirements ofthe plants Four dominant vegetation types exist in Italy as-sociated with increasing elevations from the coast to high el-evations the thermo- meso- supra- and oromediterraneanvegetation types The thermomediterranean belt is dominatedby Olea and Ceratonia at the lowest elevations the me-somediterranean belt is composed largely ofQuercus ilex-cocciferaforest supramediterranean (600ndash800 m) vegetationis defined by deciduous oak forest withQuercusandCarpi-nus and at the highest elevations the oromediterraneanbelt is characterized by anAbiesandFagus sylvaticaforest(Ozenda 1975)

The Dalmatian vegetation has been described by Horvat etal (1974) and Polunin (1980) Adriatic islands covered byQuercus ilexforest are now largely colonized by pine forestInland forests are dominated by supramediterraean vegeta-tion at higher altitudes composed of deciduous trees suchasCarpinus Ostrya FraxinusandQuercus At higher eleva-tion where summer precipitation is higher mountain forestsare mainly dominated byFagus Abies albaandPicea

Natural vegetation in the region has been affected by hu-man pressure in recent times forest clearance and inten-sive farming often result in landscapes of bush-wood oropen vegetation

4 Methods

41 Pollen record

A pollen record spanning from Last Glacial Maximumto Holocene was published by Combourieu-Nebout etal (1998) This new analysis includes 51 additional countsimproving the temporal resolution of the Holocene section ofthis core The new pollen data corresponds to time samplingintervals of from 15 to 500 yr with an average of 140 yr Sam-ples of the upper 80 cm were pollen poor and did not providereliable data possibly as a result of poor pollen preservationThe proposed pollen record from 3 to 08 m depth highlightsvegetation changes at the end of deglaciation and during mostof the Holocene fromsim 13 000 tosim 2500 cal yr BP

Pollen was obtained from marine sediments followinga previously described standard protocol (eg Faegri andIversen 1964 Combourieu-Nebout et al 1998) Samplesunderwent successive treatments with HCl and HF acidsand several sievings to obtain concentrated pollen residuesPollen grains were counted on slides with an Olympus lightmicroscope at x500 magnification Between 150 and 300grains were counted in each sample based on pollen concen-tration Pinus is over-represented in most marine samplesBecause of over-representation ofPinus in the MD 90-917pollen samples all assemblages had at least 100 non-Pinuscounts (Heusser and Balsam 1977 Turon 1984)

Pollen diagrams use sums that excludePinuspollenPinuspercentages were calculated using the total sum of identifiedgrains Pteridophyta percentages are reported using the totalsum of all identified pollen grains and Pteridophyta Pollenzones were visually determined and confirmed using clusteranalysis (CONISS software Grimm 1987) (Fig 2a b)

Fossil pollen assemblages from the MD 90-917 core showsimilarities to a range of vegetation types from semi-desert to



(b)

Fig 2 Pollen diagram of the core MD 90-917(a) arboreal taxa(b) Mediterranean and herbaceous taxa Pollen percentages are calculatedon a sum excluding Pinus Pinus percentages are calculated on the total pollen sum Pteridophyta percentages are calculated on the totalpollen sum Pollen curves are presented in calendar yr BP (correspondence between age and depth is presented on the left) On the right ofthe diagram are represented the pollen zones and results of the cluster analysis (CONISS software Grimm 1987)



Table 2Short description of the pollen zonation of the MD 90-917 pollen diagram

Pollen zone Interval cm Age kyr Pollen zone signature Main biome Climate

MD 90 917-XII 80ndash120 2641ndash3671 Pinus (50 ) darr Abies (lt 4 ) and Picea (lt 6 ) Quercus(45 )Carpinus(6 )Corylus in low percentagesuarrFagus (4 )Mediterranean taxaQ ilex (uarr 12 ) Olea (3ndash4 ) andPistacia(uarr4 )uarr Cichoriodeaeuarr Artemisia(15 )

Warm temperateMediterranean oakforest

Warm and humidstability inT

uarr

seasonalP mm

MD90 917-XI 128ndash140 3877ndash4186 Pinus (50 ) uarr Abies (lt 15 ) Picea (3 ) Quercus (40 )Carpinus (6 ) Corylus and Fagusin low percentages Mediter-ranean taxa in low percentagesQ ilex (4 )

Cool mixed forest Cool and relativelyhumid

MD90 917-X 142ndash179 4244ndash5579 Pinus low (20ndash40 ) low percentages inAbies and Picea(lt 6 ) darrQuercus (45 ) uarr Carpinus (uarr10 ) and Corylus(uarr10 ) Mediterranean taxa in stable percentages Herbs instable percentages slight increase in steppic taxa (5ndash10 )

Warm temperateoak forest

Warm and humidstability inT P mmslight decrease

MD90 917-IX 180ndash220 5623ndash6960 Pinus low (20ndash40 ) low percentages inAbies andPicea(lt 6 ) Temperate trees stableuarr Quercus (uarr45ndash65 )darr Carpi-nus (˜5 ) andCorylus (˜5 ) uarrMediterranean taxaQ ilex(6ndash7 ) Olea PhillyreaandPistacia uarr in herbs Cichoriodeae(5ndash10 ) steppic taxa low


Warm and humidstability inT andP mm slightlyreincrease

MD90 917-VIIIb 221ndash232 7010ndash7620 Pinus (20ndash50 )darr Abies (lt 10 ) and Picea (lt 5 ) decreaseuarrTemperate treesQuercus stable (lt 45 ) Corylus (5ndash20 ) Carpinus (4ndash10 )Fagus(uarr5 ) Ulmus (uarr5 ) Alnus(3ndash8 ) Mediterranean taxa stable (Q ilex mainlyOleaandPista-cia presence)darrHerbs (Asteraceaelt 5 Poaceaelt 5 ) steppictaxa low (10 )


Warm and humidincrease inT andP mm slight de-crease

MD90 917-VIIIa 233ndash234 7700ndash7779 Pinus (50 )uarr Abies(uarr25 ) andPicea(uarr11 )darr Temperatetrees Quercus (lt 45 ) Corylus Carpinus absentFagus(uarr10 ) Mediterranean taxa low Herbs low steppic taxa low (slightuarr Chenopodiaceae 2ndash5 )

Cool mixed forest Cool and humid

MD90 917-VII 235ndash238 7859ndash8058 Pinus (30ndash60 )Abies and Picealow Temperate treesQuercus(30ndash60 ) Corylus (lt 5 ) Carpinus (4 ) Fagus (uarr10 )Mediterranean taxa low (Q Ilex mainly Olea presence)darrHerbs(Asteraceaelt 5 Poaceaelt 5 ) steppic taxa nearly absent


Warm and humidMax T andP mmslight reincrease

MD90 917-VI 239ndash250 8157ndash8456 Pinus (30ndash60 )Abies and Picea low Temperate treesQuer-cus (40 )Corylus(5ndash10 )Carpinus (4ndash10 )Fagus(uarr 8 )Mediterranean taxauarr (Q ilex mainlyOleaandPistaciapresence)darr Herbs (Asteraceaelt 5 Poaceaelt 5 ) steppic taxa low


Warm and humidMax T andP mmslight decrease

MD90 917-V 252ndash258 8678ndash9445 Pinus ( around 50 )Abiesand Picea low uarr Temperate treesmaximumQuercus (45ndash70 )Corylus (5ndash10 )Carpinus (2ndash10 ) Mediterranean taxa low (Q ilex only) darr Herbs (Asteraceaelt 5 Poaceaelt 5 ) and steppic taxa low


Warm and humid ndashMax T andP mm

MD90 917-IV 259ndash269 9546ndash10737 Large decrease inPinus (lt 50 )AbiesandPicea nearly absentIncrease in temperate treesQuercus (40 ) Corylus (5ndash10 )Carpinus (2ndash4 ) Mediterranean taxa low (Q ilex and first oc-currence inOlea) darrHerbs (Asteraceaelt 12 Poaceaelt 7 )steppic taxa very low


ProgressiveuarrT andP mm

MD90 917-III 270ndash282 10858ndash11742 Pinus abundant (60ndash80 )Abiesand Picea percentages lowQuercus(20ndash40 ) darrCorylusandCarpinus slightuarrAlnus Be-tula Tilia uarr Herbs (Asteraceae 20 )uarr Chenopodiaceaegt 10

Grassland mixed toopen oak forest

Cool andmoderately dry

MD90 917-II 284ndash288 11821ndash11980 Pinus abundant (65ndash75 )Abies and Picea percentages lowuarrQuercus (15ndash40 )Corylus (uarr10 ) Carpinus (3 ) darrHerbs(Asteraceaelt 20 Poaceae around 10 )darrsteppic taxalow

Open Warm mixedforest

Slight increase inT andP mm

MD90 917-I 290ndash350 12060ndash13257 Pinusabundant (65ndash90 )Abies and Piceapercentages lowLow percentages in temperate treesQuercuslt 20 AlnusBetula Carpinuspresent Herbs abundant Asteraceae 10ndash20 Poaceae around 10 Steppic taxa abundant(Artemisia 15 Chenopodiaceae 10 Ephedraup to 7 )

Steppe Cold and dry



mountain deciduous and coniferous forest The interpretationof these assemblages follows the modern plantndashclimate rela-tionships in Eurasia and northern Africa (Woodward 1987Peyron et al 1998)

42 Pollen-based climate reconstructions

To obtain a robust quantitative reconstruction of Holoceneclimate changes in the Adriatic Sea two methods have beenused to reconstruct climate parameters modern analoguestechnique (MAT Hutson 1980 Guiot 1990) and weightedaverage partial least squares regression (WAPLS Ter Braakand Juggins 1993) Both methods have been used exten-sively for the Holocene climate reconstruction from terres-trial and marine pollen records located in the Mediterraneanregion (eg Davis et al 2003 Pross et al 2009 Peyron etal 2011 2013 Joannin et al 2012 Kotthoff et al 2008a bCombourieu-Nebout et al 2009 Dormoy et al 2009 De-sprat et al 2013 Sadori et al 2013) with results that aregenerally supported by other independent proxies (eg lakelevels alkenones charcoals etc) and other regional records

Like most classic pollen-based climate reconstructiontechniques MAT aims to quantitatively reconstruct past cli-mate from fossil assemblages based on the present-day envi-ronment MAT compares past assemblages to modern pollenassemblages without requiring real statistical calibrationusing squared-chord distance to determine the degree ofsimilarity between samples with known climate parameters(modern pollen samples) to a sample for which climate pa-rameters are to be estimated (fossil pollen sample) In con-trast to the MAT the WAPLS is a true transfer function thatrequires a statistical calibration between environmental vari-ables (climate parameters) and modern pollen assemblagesWAPLS is a common regression method which supposes thatthe relationships between pollen percentages and climate pa-rameters are unimodal More details on the MAT and on theWAPLS are given in Peyron et al (2013)

Both methods are calibrated using the high-quality andtaxonomically consistent dataset compiled by Dormoy etal (2009) This dataset contains more than 2000 samplesfrom the Mediterranean region to which were added 26modern surface samples from around Lake Trifoglietti (Cal-abria southern Italy) collected by Joannin et al (2012) andfrom around Mount Altesina and Lake Preola (Sicily)Pi-nus pollen was removed from the modern and fossil sam-ples because of its over-representation in marine pollen as-semblages In this study the quantitative climate reconstruc-tions for the core MD 90-917 were performed using an-nual (TANN) coldest month (MTCO) and warmest month(MTWA) temperature and annual (PANN) spring (Pspr)summer (Psum) autumn (Paut) and winter (Pwin) precipi-tation to emphasize changes in seasonality Results are ex-pressed as reconstructed values and anomalies (reconstructedvalues minus present-day values) (Fig 3)

43 Sediment clay fraction

Samples were prepared following standard protocols de-scribed in Bout-Roumazeilles et al (2007) All sampleswere first decalcified with 02N hydrochloric acid Excessacid was removed by repeated centrifugations The clay-sized fraction (lt 2 microm) was isolated by settling and ori-ented on glass slides (oriented mounts) Three XRD (X-ray diffraction) determinations were performed (a) untreatedsample (b) glycolated sample (after saturation for 12 h inethylene glycol) and (c) sample heated at 490C for twohours The analyses were run on a Philips PW 1710 X-ray diffractometer between 249 and 325 Each clay min-eral is characterized by its layer plus interlayer interval asrevealed by XRD analysis Smectite is characterized by apeak at 14A on the untreated sample test which expandsto 17A after saturation in ethylene glycol and retracts to10A after heating Illite presents a basal peak at 10A onthe three tests (natural glycolated and heated) Kaolin-ite is characterized by peaks at 7 and 357A on the un-treated sample and after saturation in ethylene glycol Bothpeaks disappear or are strongly reduced after heating Semi-quantitative estimation of clay mineral abundances basedon the pseudo-voigt deconvolution for the doublet kaolinite-chlorite (357Andash353A) was performed using the softwareMacDiff developed by Petschick (2001) Here we use the il-lite to kaolinite and smectite to kaolinite ratios in order toevaluate the balance between inputs from rivers respectivelyPo River (illite) and eastern Apennine rivers (smectite) ver-sus eolian inputs from the south (kaolinite) These ratios arecalculated using the semi quantitative estimation of each claymineral

5 Origin of the continental proxy data

51 Pollen inputs

Paleoenvironmental interpretation of the changes depicted inthe marine pollen assemblages is based on the assumptionthat the pollen signal recorded in the marine core reflects theregional vegetation across an area of several hundred squarekilometers (Heusser and Balsam 1977 Dupont and Wyputta2003 Hooghiemstra et al 1992 2006) Pollen compositionin deep basins of the Adriatic Sea must then be strongly af-fected by the dominant winds and hydrological currents thatbring pollen grains from their source to the site in which theyare ultimately carried

Most pollen grains within the MD 90-917 core and espe-cially arboreal pollen fromQuercus Quercus ilex CarpinusCorylus Fraxinus andAbiesare likely to originate on ei-ther coast of the Adriatic Basin given similarities in the mod-ern vegetation community structure between coasts (Ozenda1975 Horvat et al 1974) However natural stands ofOleaare nearly absent on the Dalmatian coast today and are only



MTCO degC

MTWA degC

PANN mmTANN degC

Psum mm

Pwin mm

MATWAPLS Present day value

8200

yr e

vent

Youn

ger D

ryas

-8

-4

0

4

8

2000 4000 6000 8000 10000 12000

10

15

20

25

30

2000 4000 6000 8000 10000 12000

5

10

15

20

2000 4000 6000 8000 10000 12000200

400

600

800

2000 4000 6000 8000 10000 12000

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

0

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

2000 4000 6000 8000 10000 12000Age (cal yr BP)

PB

OP

B

2000 4000 6000 8000 10000 12000Age (cal yr BP)

4200

yr

even

t

10

12

14

18

16

10

12

14

18

16

Youn

ger D

ryas

PB

OP

BSST degC (foraminifersa) SST degC (foraminifera)

8200

yr e

vent

Fig 3 Quantitative climate reconstructions from pollen analyses obtained by the MAT (red dotted line corresponds to error) and WAPLS(blue) methods Present-day values for temperature (MTCO MTWA and TANN) and precipitation (Pwin Psum and PANN) are representedby blue stars SSTs (blue) from the same core are reported at the base of the figure for comparison with the pollen inferred reconstructionAll curves are presented in cal yr BP Colored bars underline climate events YD (pink) PB (grey) PBO and Holocene events (blue)

found in southern Greece on the Balkan Peninsula (Ozenda1975) thusOleapollen grains are more likely to arrive fromthe Italian coast Nevertheless olive cultivation by human so-cieties during the last millennia (Mercuri et al 2013) maycontribute to broaderOlearepresentationPiceatoday growsonly in the Dalmatian mountains in the northern Apenninesand the Alps therefore its pollen is likely to be derived from

these areas brought to the Adriatic core site by Po River in-flow Piceamay also be carried by eastern winds howeverPicea pollen grains are considered to be transported rela-tively short distances by wind in Europe (Hicks 2001) thusPiceapollen is probably of northern origin deposited to thesite by discharge from the Po River



Ultimately pollen inputs into the Adriatic Basin will drawa broad regional picture of Italian and Balkan vegetationfrom the wide range of habitats extending from the coaststo the highlands

52 Clay mineral origin

The Po River and the eastern Apennine rivers are the mainsources of clay-sized particles in the western Adriatic Al-though Apennine rivers have relatively small drainage areasstudies have shown that they have very high sediment contri-butions relative to their drainage area (Frignani et al 1992Milliman and Syvitski 1992 Alvisi et al 1996 Bartolini etal 1996 Sorgente 1999 Tomadin 2000) Detrital materialin the Po River comes largely from the Alps supplying illiteassociated with chlorite to the Adriatic Sea (Chamley 1989Alonso and Maldonado 1990 Tomadin 2000) Because il-lite is resistant to degradation and transport it generally repre-sents the relative contribution of physical weathering to sed-imentation As a result illite dominates the clay mineral sed-iment fraction in the deeper parts of the Adriatic (Tomadin2000) By contrast Apennine sediment sources are rich insmectite which is mainly dispersed southeastward along thecoast with further downslope transport toward the deep basinthrough seasonal gradient and turbidity currents (Franco etal 1982 Tomadin 2000)

Croatian rivers to the east containing illite and kaolinitewith minor smectite components (Durn et al 1999) and Al-banian rivers to the south carrying smectite supply sedimentto the eastern Adriatic The contribution of these river inputsis reduced because of very low riverine terrigenous loads andparticles trapped along the Adriatic margin (eg Tomadin2000 Cattaneo et al 2003)

The eolian contribution to deep-sea sediments is of majorimportance in the Mediterranean Massive desert dust plumesexport clay-mineral particles toward the Mediterranean (Reaet al 1985 Guerzoni and Chester 1996)via regional merid-ian wind systems (Prospero 1981 Loye-Pilot et al 1986Pye 1987 Bergametti et al 1989 Tomadin and Lenaz1989 Guerzoni and Chester 1996 Guerzoni et al 1999Moulin et al 1997 Rodriguez et al 2001 Torres-Padron etal 2002 Ginoux et al 2004) Illite and kaolinite can also betransported through eolian processes Kaolinite abundance ishighest in the southeastern Sahara compared to western Sa-hara while illite displays the reverse pattern (Avila et al1997 Caquineau et al 1998 Guerzoni et al 1999) Thesoutherly sirocco wind most likely supplies eolian kaolinite-rich dusts from North Africa to the Adriatic Sea during springand summer (Goudie and Middleton 2001)

6 Vegetation and climate for core MD 90-917

Palynological investigations using marine sediment coreshave confirmed close links between pollen data from marine

sediments and regional vegetation and the ability of marinepalynology to reconstruct climate changes at a regional scale(eg Heusser and Balsam 1977 Hooghiemstra et al 19922006 Combourieu-Nebout et al 2009) The reconstructedvegetation history for MD90-917 is similar to other Italianand Balkan sequences for the Holocene which suggests thepollen from MD 90-917 can reveal the consequences of theregional climatic change during the last 13 000 cal yr BP oncentral Mediterranean vegetation (Table 2 Fig 2a b)

The Younger Dryas event (GS-1 event ndash YD) is charac-terized by the dominance of semi-desert elements such asArtemisia Chenopodiaceae andEphedra(Fig 2b) accom-panied by Poaceae and Asteraceae which developed ex-tensively around a lowered Adriatic Sea (also described inCombourieu-Nebout et al 1998) This steppic vegetationis recorded in numerous regional pollen records (eg Wattset al 1996 Magri and Sadori 1999 Denefle et al 2000Allen et al 2002 Bordon et al 2009 Kotthoff et al 2011Combourieu-Nebout et al 2009 Fletcher et al 2010 De-sprat et al 2013) Steppic pollen taxa indicate that cold-arid climate conditions prevailed over the whole Mediter-ranean Basin and especially in the Adriatic Basin during theYounger Dryas (Figs 3 4)

Between 12 000 and 11 700 cal yr BP semi-desert ele-ments are replaced by arboreal associations composed byQuercus Carpinus CorylusandAbiesshowing the afforesta-tion driven by climatic warming at the beginning of theHolocene (Fig 2a) Although it begins at an older age in theMD 90-917 marine core such a spread of temperate taxa maybe related to what is observed in several pollen sequencesacross the central Mediterranean (eg Watts et al 1996Sadori and Narcisi 2001 Allen et al 2002 Sadori et al2011 Di Rita et al 2013) and corresponds to the Preboreal(PB) improvement at the Younger DryasHolocene transi-tion (Bjork et al 1996) Forest expansion is interrupted by astrong drop in arboreal taxa around 11700 cal yr BP markedby an increase in herbaceous taxa (Asteroideae and Cichori-odeae) and semi-desert taxa excluding Artemisia (Fig 2ab) Climate reconstructions using pollen data show a de-crease in temperature and precipitation at this time for thetwo methods (Figs 3 4) Although lake records from thecentral Mediterranean are rare the same pattern can be seenan increase in temperate forest followed by a slight returnof herbs (eg Lake Trifoglietti ndash Joannin et al 2012 Valledi Castiglione ndash Di Rita et al 2013) Thus the climate os-cillation recorded in the MD90-917 appears to follow thepattern observed in the Greenland ice coresrsquoδ18O record(Fig 5 Rasmussen et al 2007) and probably correspondsto the sim 11 300 cal yr BP Preboreal oscillation recorded inthe Greenland ice cores (PBO Bjork et al 2001 Magnyet al 2012)

In the early Holocene the general trend of the vegeta-tion data and pollen-based climate reconstructions precedethe foraminifer-based SST (sea surface temperature) from thesame core (Siani et al 2013) (Figs 3 5) in the same pattern



0

10

20

30

40

50

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

0

10

20

30

400

500

600

700

800

900

1000

300

100

200100

200

100

200

300

100

200

60

80

70

0

10

20

30

40

50

60

80

70

0

10

20

0

10

20

30

0

10

20

30

40

0

10

Alti

tudi

nal

taxa

()

Oth

er te

mpe

rate

Cor

ylus

()

Tem

pera

te fo

rest

d

ecid

uous

Q

uerc

us (

)Med

iterr

anea

nta

xa (

)S

emi-d

eser

tta

xa (

)P

sum

(mm

)

Her

bace

ous

taxa

()

Psp

r (m

m)

Pau

t (m

m)

PAN

N (m

m)

Pw

in (m

m)

Mea

n te

mpe

ratu

re (deg

C)MTWA

MTCO

TANN

Age (cal yr BP)

Age (cal yr BP)

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

917-V

917-X

II

917-X

I

917-X

917-I

X

917-V

IIIb

917-V

I

917-I

V

917-I

II

917-I

917-I

I

917-V

IIIa

917-V

II

Pol

len

zona

tion

Present-day value

Fig 4 Selected pollen curves of core MD 90-917 and pollen based reconstruction of precipitation (seasonal Pspr Psum Paut Pwin andannual PANN) and temperature ( MTCO MTWA and TANN) All curves are presented in cal yr BP Grey curves are the 3 pts-smoothedcurves Colored stars represent the present-day values for each climate parameter



-40

-34

0

20

40

60

80

2000 4000 6000 8000 10000 12000

0

60

40

20

2000 4000 6000 8000 10000 12000

Age (cal yr BP)

NGRIP δ18O (permil)

Age (cal yr BP)

Semi desert ()

Temperate forest ()

-36

-38

-42

82

00 y

r eve

nt

PB

O

Youn

ger

Dry

as

PB

18

16

14

12

10

SST (degC foraminifera)

Fig 5 Comparison between two main pollen groups (this paper)temperate trees (green) semi-desert (red) and SST (blue) from thecore MD 90-917 and NGRIP oxygen isotope record (NGRIP mem-bers 2004 Lemieux Dudon 2011) Colored bars underline climateevents YD (pink) PB (grey) PBO and 8200 cal yr BP event (pur-ple) Curves are plotted according to their respective age models incal year BP The time lag observed between MD 90-917 and NGRIPrecords during the YDndashHolocene transition is probably due to14Cage uncertainties displayed in the MD 90-917 age model The ques-tion marks indicate potential events not sufficiently documented inthe record

as the alkenone-inferred SST response (Sicre et al 2013) Toexplain such discrepancies Sicre et al (2013) suggest a localresponse in alkenone-inferred SST record In fact the tem-poral lead shown in MD 90-917 vegetation and pollen-basedclimate reconstruction records relative to foraminifer-basedSST should not be related to local behavior as it would implya time lag with other Italian records such as Lake Trifogli-etti (Joannin et al 2012) and Lake Vico (di Rita et al 2013)records which would be amazing Of course the MD 90-917 record represents the regional vegetation changes of thecentral Mediterranean and thus integrates vegetation changesevidenced in these contemporary Italian records Another ex-planation might be that the lead in vegetation response rela-

tive to foraminifer inferred SST from the same core suggeststhat the vegetation response to the Preboreal oscillation inthe region could be driven by precipitation changes ratherthan temperature variations which is quite different to theclassic feature of climate changes observed at the beginningof interglacials (temperature rise generally preceding precip-itation increase) Thus the discrepancy between the differentproxies obtained in the core MD 90-917 can be related to thesampling resolutions which are not strictly similar in thispart of the core and the delay of SST response compared tovegetation may be due to the lack of samples in the PBndashPBOinterval Given this age discrepancies between continentalice records and the MD 90-917 marine core vegetation proxymay be due to14C age control uncertainties in this part ofthe core and that the vegetation event at 11 700 cal yr in MD90-917 may be related without doubt to the Preboreal event

After 11 000 cal yr BP a mixed deciduous forest (mainlydominated byQuercuswith regular occurrence of and ac-companied byCorylus Carpinus Fagus Alnus Betula) ex-pands (Fig 2a b) The mixed forest peaks around 7000 cal yrBP corresponding to the ldquoHolocene climate optimumrdquo De-ciduous forest expansion occurs in Italy and in the BalkanPeninsula at the same time (eg Watts et al 1996 Jahnsand van der Bogaard 1998 Denefle et al 2000 Allen etal 2002) Evidence for a climate optimum at 7000 cal yr BPagrees well withδ18O from planktonic foraminifers (Fig 6)and the foraminifer inferred SST increase from the MD 90-917 core (Siani et al 2013) Increases in regional forestbiomass also concur withδ13C depletion in the MD 90-917sediments linked to sea-level rise (Siani et al 2013) in theearly-Holocene since depletion inδ13C in core sedimentsmay also be interpreted as an intensification of nutrient sup-ply from the continent to the Adriatic Sea

Although temperate forest dominates the vegetation recordfrom 11 000 untilsim 7000 cal yr BP there are several com-positional shifts in that time (Figs 2a 4) Inferred vegetationis dominated byQuercus which declines repeatedlyQuer-cusis sometimes replaced by a mixed deciduous forest withCorylus Carpinus andFagus occasionally associated withFraxinusandAlnusor is replaced by a mixed coniferous for-est withAbiesandPicea In the first caseQuercusdominatedforests are replaced with mixed deciduous forests at around8300 and 7500 cal yr BP possibly the expression of a for-est opening resulting in the expansion of heliophilous taxasporadically developing in theQuercus forest andor on theforest edges Such changes might be due to slight changesin summer temperature and precipitation (Fig 3) or changesin the spring and autumn (Fig 4) In the second caseQuer-cus forest is replaced by mixed coniferous forest at 7700cal yr BP withAbiesandPicea This change might reflect aslight cooling andor change in the seasonality of precipita-tion particularly in the Balkan massifs andor on the northernpart of Adriatic Basin given that delivery ofPiceapollen tothe core occurs (i) from the rivers of the northern Adriaticslopes andor (ii) from northern Italy and Alps by Po River



- 250

- 150

- 050

20

40

60

80

100

2000 4000 6000 8000 10000 12000 14000

050

150

250

Trees

18δ O permil

13Cδ C permil

Age (cal yr BP)

18

810121416

SST degC (planctonic foraminifera)

Fig 6 Comparison of the total tree percentage curve (green) (thispaper) and theδ18O (blue) δ13 C (red) and SST (C purple)records from the MD 90-917 core (Siani et al 2013) All curvesare presented in cal yr BP

discharges and then by marine flowsPicea has not beenobserved in central and southern Italy during the Holoceneonly occurring in the Po Valley northern Italy and over thenorthern Balkans (Ravazzi et al 2002 Van der Knaap et al2005)AbiesandPiceapollen grains are not well-transportedby wind (only at 5 m away from the trees) and rarely dom-inate airborne pollen assemblages (Hicks 2001 Sjogren etal 2008) As a result theAbiesandPiceaincrease atsim 7700cal yr BP could therefore express the development of mixedAbiesandPiceaforests in the north and northeast mountainsaround the Adriatic Basin with the pollen being transportedthrough river flows

Temperate forest dominated byQuercushave fluctuatedbetween 7000 and 5000 cal yr BP decreasing after 4500cal yr BPQuercusbecame less abundant whileCarpinusCorylusandAlnus increased This marks progressive open-ing of the forest corresponding to decreases in summer pre-cipitation reflected in the pollen-based climate reconstruc-tions based on the MAT and the WAPLS (Fig 3) A coldmoist event is inferred from the rise inAbies Fagus and to alesser extentPiceaaround 4000 cal yr BP (Fig 4) Increasedprecipitation would have increased river flows bringing thepoorly dispersedAbiespollen to the core location The up-per samples of the core sequence to 3000 cal yr BP show anincrease in herbaceous taxa (Asteraceae and Poaceae) con-tinuing the trend of declining tree cover probably influencedby increasing of human impacts in coastal Italy and BalkanThe increase in steppic elements such asArtemisia duringthis period may also indicate regional drying particularly inthe summer and may represent the establishment of the sum-

mer conditions associated with the modern Mediterraneanclimate

Mediterranean taxa primarilyQ ilex are continuouslypresent in the pollen record beginning at 8000 cal yr BPThe Mediterranean forest begins to diversify after 6000 calyr BP with the continuous presence ofOlea and Phillyreapollen and after 3500 cal yr BP with the presence ofPista-cia This coincides with the decreasing abundance of decid-uous trees especially deciduousQuercusand the increasein herbs such as Asteraceae and steppic taxa (Artemisiaex-cluded) at 3500 cal yr BP This pattern has been observedelsewhere in the central Mediterranean (Sadori et al 2011)and marks a shift in the precipitation regime with decreas-ing summer precipitation and a progressive establishment ofthe modern Mediterranean climate (Figs 3 4)Oleathe em-blematic plant of the Mediterranean area occurs weakly atthe beginning of the Holocene and remains sporadic up to6000 cal yr BP generally associated only with warming cli-mate After 6000 cal yr BPOlea is continuously present butis most abundant in the upper part of the record around 3000cal yr BP in agreement with the central Mediterranean veg-etation history proposed by Sadori et al (2011) using conti-nental recordsOlea is present in marine records across theMediterranean at this time (eg Combourieu-Nebout et al2009 Desprat et al 2013) and may also reflect the develop-ment of olive cultivation at Italian sites (eg Lake PergusaMercuri et al 2013 Sadori et al 2013) and possibly in theBalkans

7 Climate interpretation

71 Temperature pattern

Temperature estimates from the MD 90-917 pollen sequence(Figs 3 4) are the first to be reconstructed from Adriaticmarine sediments These results can be compared to temper-ature signals reconstructed from terrestrial pollen sequencesat lakes Ledro Trifoglietti Accesa and Pergusa (Fig 1)for a direct landndashsea comparison (Peyron et al 2013) Wepresent TANN MTCO and MTWA temperature reconstruc-tions from MD 90-917 provided by two methods the MATand the WAPLS (Fig 3) The lowest MTCO in the record oc-curs before 12 000 cal yr BP when temperatures are 5ndash7Clower than preceding and subsequent periods After thesecoldest values MTCO increases towards higher values andthen oscillate around 0C with a slight decreasing trend Thetemperatures obtained with the WAPLS were lower duringthe YD than during the Holocene which makes sense andsuggests that the MAT may provide lower quality results dur-ing the Younger Dryas and Preboreal (Ortu et al 2010ab) However Holocene results are consistent between meth-ods Temperatures (TANN and MTWA) minima are observedaround 7700 cal yr BP at which point they increase graduallythrough the upper part of the record (Fig 3)



Temperature anomalies reconstructed between 6000 cal yrBP and present-day are similar to those proposed for south-ern Italy by Wu et al (2007) reinforcing the reliability of ourclimate reconstruction trends The reconstructed temperatureanomaly (T C reconstructed ndashT C present-day) record fromMD 90-917 also mirrors the southwestern European temper-ature curve reconstructed using more than 500 continentalpollen records by Davis et al (2003) (Fig 7) who also in-dicate low temperature anomalies during the mid-HoloceneThis confirms that the pollen record in the Adriatic reflectssouthwestern Mediterranean climate influences more so thansoutheastern influences The low anomalies likely reflect thestrength of the northndashsouth Mediterranean gradient (Magnyet al 2012) and an increase in the influence of westerlies inthe northcentral Mediterranean during the late Holocene

Temperature reconstructions indicate several cold andordry events during the last 13 000 cal yr BP the YoungerDryas event the Preboreal oscillation the 8200 cal yr BPevent and others events at 7700 7000 6400 5000 4200and 3000 cal yr BP (Fig 3) These cold events are asso-ciated with declines inQuercuspollen and have been de-tected in other marine cores from the Mediterranean (Fig 8eg Combourieu-Nebout et al 2009 Schmiedl et al 2010Fletcher et al 2010 2012 Desprat et al 2013) and maybe linked to the millennial-scale cold events recorded in theNorth Atlantic (Bond et al 1997 2001 Kotthoff et al2008a Combourieu-Nebout et al 2009 Schmiedl et al2010 Fletcher et al 2010 2012 Fletcher and Zielhofer2011 Desprat et al 2013 Magny et al 2013) The rapid re-sponse of the central Mediterranean to millennial-scale vari-ability confirms the influence of midlatitude atmospheric cir-culation in propagating changes depending on insolation andice sheet volume (Desprat et al 2013) Our reconstructionsuggests that the first two events (Younger Dryas and Pre-boreal oscillation) were characterized by sharp seasonal con-trast with respectively coldcool winters and warm summer)this pattern differs strongly from the modern climate whichestablished itself progressively after 7700 cal yr BP withwarmer winters more favorable to Mediterranean forest com-munities (Fig 3)

72 Precipitation pattern

Pollen-inferred annual precipitation for MD 90-917 showsan increasing trend from 12 000 to 7000 cal yr BP reachinghighest values between 8000 and 7500 cal yr BP (Figs 34) This trend is mirrored by Pspr and Paut precipitationbut while Psum and PANN precipitation show similar trendsof increasing precipitation from the early Holocene to 7500cal yr BP summer precipitation declines more strongly thanannual precipitation towards 3000 cal yr BP The MAT andthe WAPLS results show an unambiguous increase in sum-mer precipitation during the mid-Holocene (Fig 3) even ifthe MAT may slightly overestimate the summer precipitationand underestimate the winter precipitation (Combourieu-

1000 3000 1100050002000 4000 7000 80000 100009000 120006000

-4

-6

-8

-10

-12

2

0

-2

-4

-6MTW

A an

omal

ies

MTC

O a

nom

alie

s

MD 90-917

SW Europe

SW Europe

SE Europe

SE Europe

MD 90-917

1000 3000 1100050002000 4000 7000 80000 100009000 120006000

1

0

-1

-2

-3

2

1

0

-1

-2

0

-1

-2

-3

-4

3

2

1

0

Age (cal yr BP)

Age (cal yr BP)

MTW

A anomalies

MTC

O anom

aliesM

TWA anom

aliesM

TCO

anomalies

Fig 7 Comparison between MD 90-917 pollen-based MTCO andMTWA temperature reconstructions (MAT in black WAPLS ingrey) and the combined climate records of southeastern and south-western Europe (from Davis et al 2003) Curves are presented intemperature anomalies (reconstructed values minus present-day val-ues)

Nebout et al 2009) The Holocenersquos summer precipitationtrend inferred from our pollen data is similar to speleothemrecords in central Italy (Zhorniak et al 2011) and the east-ern Mediterranean (Bar-Matthews et al 1998 Bar-Matthewsand Ayalon 2011) although speleothem-record featuresshould be linked to changes in winter precipitation Howeverinterpretations of speleothem records in terms of precipita-tion seasonality in the central and eastern Mediterranean re-main a matter of debate (Zhorniak 2011) (Fig 9)

The summer precipitation record also matches the lake-levels pattern described by Magny et al (2012) which sug-gests that lakes located southward of 40 N experienced highlake levels during precipitation maxima in the early to mid-Holocene in Italy while northern Italian lakes were char-acterized by low lake levels The summer precipitation in-ferred from four Italian pollen records appears to confirmthe opposing precipitation regimes for northern and southernsites in the Mediterranean during the Holocene (Peyron etal 2013) There is strong agreement between the high lake



Combourieu-

YD YD

SL 152

Schmiedl et al (2010)

Kotthoff et al(2008a)

200 40 60

Quercus ()

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age

(cal

yr B

P)

ACP2

ACP3

ACP4

ACP5

ACP6

ACP7

ACP8ACP9

YD

NAC1

NAC2

NAC3

NAC4

NAC5

NAC6

NAC7

NAC8

NAYD

MD 90-917

(this paper)Bond et al

(2001)

North Atlantic Mediterranean

cold events

YD

West-CentralEurope

YD

AdC1

AdC2

AdC4

AdC6

AdC7

AdC8

AdC3

AdC5

MD 01-2797

Desprat et al(2013)

MD 99-2043

Fletcher et al(2010 2012)

Magny et al(2012)

ODP 976

Nebout et al(2009)

60

Lake-levelhighstands

PBO

BO

8200

9200

4200

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age (cal yr BP)

Fig 8 Comparison of millennial cold events recorded in MD 90-917 with the cold events recorded from vegetation changes in otherMediterranean marine cores (Kotthoff et al 2008a Schmiedl et al 2010 Fletcher et al 2010 2012 Combourieu-Nebout et al 2009Desprat et al 2013) On the left are the North Atlantic Bond events (Bond et al 2001) and on the right the periods of high lake levelsrecorded in the southern Alps (Magny et al 2012)

levels reconstructed in Sicily the wet summer conditions re-constructed for southern Italy by Peyron et al (2013) and thesummer precipitation inferred from the MD 90-917 pollenrecord The MD 90-917 summer precipitation curve showssimilarities to the northern Italy Lake Ledro record duringthe early Holocene while the mid-to-late Holocene precip-itation curve fits with the southern Mediterranean summerprecipitation trend (Fig 8 Peyron et al 2013) The MD 90-917 precipitation pattern may support at the mid-Holocene acorrelative increase in winter and summer precipitations withhigh summer precipitation and cooler temperatures at time ofthe boreal insolation maximum Such a signal may illustratethe possible conflict of mid-latitudinal and subtropical mon-soonal climatic systems in the Mediterranean area (Tzedakis2007 Revel et al 2010 Desprat et al 2013 Magny et al2013) and surely emphasizes the fact that our record is lo-cated at the confluence between northern and southern in-fluences Climate changes recorded here are clearly drivenby insolation changes and likely reflect the increasing influ-ence of westerlies during the second half of the Holocene asdemonstrated by the rise in winter precipitation in the upperpart of our record

Summer precipitation shows good fit to (i) sea surfacesalinity changes from the same core interpreted as increas-ing runoff into the Adriatic deep sea basin (Siani et al 2013)and (ii) changes in clay mineral composition (Fig 9) Claymineral records presented as illitekaolinite (IK) and smec-

titekaolinite (SK) ratios (Fig 9) reflect contributions fromthe eastern ApenninePo Rivers and dust input from Africarespectively These ratios clearly show a change in detritalinput during the late-glacial and Holocene Wind-blown dust(high kaolinite) supplied from the south by sirocco winds ismore abundant during the first part of the Holocene while therise of smectite after 9000 cal yr BP indicates the increasingcontribution from Italian rivers to the basin

Both IK and SK display a major peak at 12 500 cal yrBP (Fig 9) associated with an early rise in summer pre-cipitation suggesting important discharges from both the Poand the eastern Apennine rivers A single SK peak at 12 000cal yr BP suggests a major discharge from eastern Apen-nine rivers that is synchronous with the precipitation max-ima likely related to the Preboreal event

Annual precipitation inferred from pollen data shows twomaxima in the middle Holocene During these two eventsprecipitation seasonality and temperature display two differ-ent patterns (i) at 7700 cal yr BP a longer humid seasonis combined with low summer and winter temperatures and(ii) at 7000 cal yr BP precipitation is equitable with no no-ticeable change in temperature These two peaks occur atthe same time that clay mineral composition changes withtwo different signatures indicative of two distinct sourcesThe 7700 cal yr BP event shows a major increase in theillite kaolinite ratio indicating major discharges from thePo River This pattern of increasing inflow from the Po River



2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

100

200Psum (mm)

S1aS1b

Age (cal yr BP)

S Adriatic 18δ Ow (permil vsSMOW)

+

-

2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

Age (cal yr BP)

-6

-5 Soreq Cave δ O (permil)18

IK

Po

Riv

er

disc

harg

e

+

-

loca

l riv

ers

dis

char

ge

-

+

salin

e

+

- Sum

mer

Pre

cipi

tatio

n

-52

-49

-46

-43

-40

-37

15

20

1

2

3

Renella Cave 18δ O (permil) +

- Pre

cipi

tatio

n

+

- Pre

cipi

tatio

n

3

5

7

9

Pwin (mm)100

200 +

- Win

ter

Pre

cipi

tatio

n

SK

Lake-level highstandswest-central Europe

PBOBO82 92 42

M

D

90

917

Fig 9 Comparison of the Psum (yellow) and Pwin (red) precipita-tion changes reconstructed from MD 90-917 pollen data (MAT) andfrom bottom to top lake level highstands in western Europe (Magnyet al 2012 column) sea surface salinity reconstructed from MD90-917δ18O record (Siani et al 2013 blue) IK (pink) and SK(green) ratios from MD 90-917 andδ18O speleothem records fromthe Renella cave (Italy) (Zhornyak et al 2011 purple) and Soreqcave (Israel) (Bar-Matthews et al 1998 Bar-Matthews and Ayalon2011 grey) For MD 90-917 records thin light curves represent theoriginal data and accentuated curves are the 3 pts-smoothed curves

agrees well with increases inPiceaandAbiespollen in thesediments and confirms the role of Po River discharge in thepollen signal as well as its probable influence on declines insea surface salinity during this event (Fig 9) Both the ori-gin and discharge rate of terrigenous inputs in the AdriaticSea are seasonally modulated by precipitation distribution(Wheatcroft et al 2001 Poulain 2001) Modern seasonalprecipitation influences discharge from the northern Adriaticas is shown for recent floods in the Po River basin (Zanchet-

tin et al 2008) Therefore increases in precipitation season-ality probably occurred during the 7700 cal yr BP event par-ticularly in the northern and eastern Adriatic leading to en-hanced Po River discharges that would have driven salinitychanges in the Adriatic Sea surface and transportedPiceapollen toward the corersquos site Such a process should explainwhy vegetation change from 8200 to 7500 cal yr BP reportedfrom Trifoglietti (Joannin et al 2012) is ascribed to a dryand cold phase This dry phase does not appear in MD 90-917 which shows changes similar to those observed in otherMediterranean records (eg Sadori and Narcisi 2001 Allenet al 2002)

The second climatic event around 7000 cal yr BP showsincreasing annual precipitation without increasingPicea Atthis time sea surface waters are less saline based onδ18Ofrom MD 90-917 (Siani et al 2013) The 7000 cal yr BPevent occurs after the main peak of the illitekaolinite ra-tio indicating a reduced role for Po River discharges butthe event is synchronous with an increase in the smec-titekaolinite ratio Sedimentological studies suggest that therespective contribution of the Po River and eastern Apenninerivers has varied through time (Cattaneo et al 2003) Smec-tite is not presently supplied to the Adriatic through the PoRiver originating from Italyrsquos slopes through the small butactive eastern Apennine rivers (Fig 7) Based on the depth atwhich the core was obtained sedimentation should be mostlyunder the influence of the illite-rich Po River plume whichflows southeastward in the open sea The smectite-rich south-eastward Apennine plumes are generally restricted to shal-lower depths along the coast Smectite is transferred towardthe deep basin through turbidity and gravity currents whenriver discharge increases seasonally (Tomadin 2000) As aconsequence the 7000 cal yr BP event likely reflects a localpattern of change reflecting increases in precipitation affect-ing the central Italian coast nearest the corersquos site This sug-gests that decreases in salinity were caused by inflow fromlocal rivers These inputs persisted throughout the secondhalf of Holocene but it is likely that they were then too weakto induce any significant decrease in sea surface salinity inthe Adriatic Sea

8 Conclusions

The MD 90-917 marine pollen record illustrates the regionalresponse of central Mediterranean vegetation to long-termand millennial climate fluctuations during the late-glacial andHolocene and gives an integrated signal of temperature andprecipitation changes during that period

The early Holocene is clearly expressed in vegetationchanges and despite a short chronological time-lag withthe ice core chronology the Preboreal climate oscilla-tion is recorded in the core by an expansion of cen-tral Mediterranean forest taxa followed by the return ofopen vegetation Shifts in vegetation during the Preboreal



oscillation are probably driven by precipitation changes es-pecially summer precipitation and are correlated with inputsfrom local and regional rivers since vegetation changes oc-curred prior to SST increases recorded in the MD 90-917core

Pollen-based temperature reconstructions follow patternsreconstructed using other proxies in the southern Mediter-ranean with an increasing west and northwest climatic influ-ence Several cold events are recorded and may be linked tomillennial-scale variability during the Holocene

Precipitation reconstructed from the pollen record showsa trend of increasing annual and seasonal (winter) precip-itation during the Holocene Summer precipitation reflectsa trend seen elsewhere in the southern Mediterranean withmaximum values (around 180ndash200 mm from the two meth-ods) around 7000 cal yr BP and then declining until 3000 calyr BP towards nearly present-day values (lt 100 mm) Twosignificant peaks in precipitation occur at 7700 cal yr BP andbetween 7500 and 7000 cal yr BP that are associated withpeak runoff from the Po River and from local central Adri-atic rivers respectively

AcknowledgementsThis study is a part of the LAMA ANRprojects (MSHE Ledoux USR 3124 CNRS) financially supportedby the French CNRS (National Centre for Scientific Research)The authors received funding from French CNRS and CEA fortheir research S Goring is funded through the NSF-Macrosystemsprogram grant EF1065656 This research has been supportedby the Institut Paul Emile Victor for the Prometee cruise of theRV Marion Dufresne Palynological samples were processed byJ P Cazet and I Dormoy This is LSCE contribution no 4587

Edited by M-F Loutre

The publication of this articleis financed by CNRS-INSU

References

Allen J R M Watts W A McGee E and Huntley B Holoceneenvironmental variabilityndashthe record from Lago Grande di Mon-ticchio Italy Quatern Int 88 69ndash80 2002

Alonso B and Maldonado A Late Quaternary sedimentationpatterns of the Ebro turbidite systems (northwestern Mediter-ranean) Two styles of deep-sea deposition Mar Geol 95 353ndash377 1990

Alvisi F Beks J Frignani M Langone L Moodley L Mow-bray S Price N B and Ravaioli M Toward a sedimentand heavy metal mass balance for the western Adriatic Sea 2ndWorkshop MYP-MAST Iraklio Greece extended abstracts 59ndash64 1996

Avila A Queralt-Mitjans I and Alarcon M Mineralogical com-position of African dust delivered by red rains over northeasternSpain J Geophys Res 102 21977ndash21996 1997

Bar-Matthews M and Ayalon A Mid-Holocene climate vari-ations revealed by high-resolution speleothem records fromSoreq Cave Israel and their correlations with cultural changesHolocene 21 163ndash172 2011

Bar-Matthews M Ayalon A and Kaufman A Middle to lateHolocene (6500 yr period) paleoclimate in the Eastern Mediter-ranean region from stable isotopic composition of speleothemsfrom Soreq Cave Israel in Environment and society in timesof climate change edited by Issar A and Brown N KluwerAcademic Dordrecht 203ndash214 1998

Bartolini C Caputo R and Pieri M Pliocene-Quaternary sed-imentation in the Northern Apennines Foredeep and related de-nudation Geol Mag 133 255ndash273 1996

Bergametti G Gomes L Remoudaki E Desbois M Martin Dand Buat-Menard P Present transport and deposition patternsof African dusts to the North-Western Mediterranean in Pale-oclimatology and Paleometeorology Modern and past patternsof global atmospheric transport edited by Leinen M and Sarn-thein M Dordrecht Kluwer Academic 282 227ndash251 1989

Bjorck S Kromer B Johnsen S Bennike O HammarlundD Lemdahl G Possnert G Rasmussen TL WohlfarthB Hammer C U and Spurk M Synchronized terrestrial-atmospheric deglacial records around the North Atlantic Sci-ence 274 1155ndash1160 1996

Bjorck S Muscheler R Kromer B Andresen C S HeinemeierJ Johnsen S J Conley D Koc N Spurk M and Veski SHigh-resolution analyses of an early Holocene climate event mayimply decreased solar forcing as an important climate triggerGeology 29 1107ndash1110 2001

Bond G Showers W Cheseby M Lotti R Almasi P de-Menocal P Priore P Cullen H Hajdas I and Bonani GA pervasive millennial-scale cycle in North Atlantic Holoceneand Glacial Climates Science 278 1257ndash1266 1997

Bond G Kromer B Beer J Muscheler R Evans M N Show-ers W Hoffman S Lotti-Bond R Hajdas I and Bonani GPersistent solar influence on North Atlantic climate during theHolocene Science 294 2130ndash2136 2001

Bordon A Peyron O Lezine A Brewer S and Fouache EPollen-inferred Late-Glacial and Holocene climate in southernBalkan (Lake Maliq) Quatern Int 200 19ndash30 2009

Bout-Roumazeilles V Nebout N C Peyron O Cortijo ELandais A and Masson-Delmotte V Connection betweenSouth Mediterranean climate and North African atmospheric cir-culation during the last 50000 yr BP North Atlantic cold eventsQuaternary Sci Rev 26 3197ndash3215 2007

Caquineau S Gaudichet A Gomes L Magonthier M-C andChatenet B Saharan dust clay ratio as a relevant tracer to assessthe origin of soil-derived aerosols Geophys Res Lett 25 983ndash986 1998

Cattaneo A Correggiari A Langone L and Trincardi F Thelate-Holocene Gargano subaqueous delta Adriatic shelf Sedi-ment pathways and supply fluctuations Mar Geol 193 61ndash912003

Chamley H Clay sedimentology Berlin Springer-Verlag 623 pp1989



Combourieu-Nebout N Paterne M Turon J-L and Siani GA high-resolution record of the Last Deglaciation in the Cen-tral Mediterranean sea palaeovegetation and palaeohydrologicalevolution Quaternary Sci Rev 17 303ndash332 1998

Combourieu-Nebout N Peyron O Dormoy I Desprat S Beau-douin C Kotthoff U and Marret F Rapid climatic variabilityin the west Mediterranean during the last 25 000 years from highresolution pollen data Clim Past 5 503ndash521 doi105194cp-5-503-2009 2009

Davis B A S Brewer S Stevenson A C Guiot J and datacontributor The temperature of Europe during the Holocene re-constructed from pollen data Quaternary Sci Rev 22 1701ndash1716 2003

Denefle M Lezine A M M Fouache E and Dufaure J J A12000-year pollen record from Lake Maliq Albania QuaternaryRes 54 423ndash432 2000

Desprat S Combourieu-Nebout N Essallami L Sicre M ADormoy I Peyron O Siani G Bout Roumazeilles V andTuron J L Deglacial and Holocene vegetation and climaticchanges in the southern Central Mediterranean from a directland-sea correlation Clim Past 9 767ndash787 doi105194cp-9-767-2013 2013

Di Donato V Esposito P Russo-Ermolli E Scarano A andCheddadi R Coupled atmospheric and marine palaeoclimaticreconstruction for the last 35 ka in the Sele Plain-Gulf of Salernoarea (southern Italy) Quaternary Int 190 146ndash157 2008

Di Rita F and Magri D Holocene drought deforestation and ev-ergreen vegetation development in the Central Mediterranean a5500 years record from Lago Alimini Piccolo Apulia SoutheastItaly Holocene 19 295ndash306 2009

Di Rita F Anzidei A P and Magri D A Late glacial and earlyHolocene pollen record from Valle di Castiglione (Rome) Veg-etation dynamics and climate implications Quaternary Int 28873ndash80 2013

Dormoy I Peyron O Combourieu-Nebout N Goring S Kot-thoff U Magny M and Pross J Terrestrial climate variabil-ity and seasonality changes in the Mediterranean region between15 000 and 4000 years BP deduced from marine pollen recordsClim Past 5 615ndash632 doi105194cp-5-615-2009 2009

Drescher-Schneider R de Beaulieu J-L Magny M Walter-Simonet A-V Bossuet G Millet L Brugiapaglia E L andDrescher A Vegetation history climate and human impact overthe last 15 000 years at Lago dellrsquoAccesa (Tuscany CentralItaly) Veg Hist Archaeobot 16 279ndash299 2007

Dupont L M and Wyputta U Reconstructing pathways of aeolianpollen transport to the marine sediments along the coastline ofSW Africa Quatern Sci Rev 22 157ndash174 doi101016S0277-3791(02)00032-X 2003

Durn G Ottner F and Slovenec D Mineralogical and geochem-ical indicators of the polygenetic nature of terra rossa in IstriaCroatia Geoderma 91 125ndash150 1999

Faegri K and Iversen J Textbook of Pollen Analysis Munks-gaard Copenhagen Denmark 237 pp 1964

Finne M Holmgren K Sundqvist H S Weiberg E and Lind-blom M Climate in the eastern Mediterranean and adjacentregions during the past 6000 years ndash A review J Archaeol Sci38 3153ndash3173 doi101016jjas201105007 2011

Fletcher W J and Zielhofer C Fragility of Western Mediterraneanlandscapes during Holocene Rapid Climate Changes Catena

103 16ndash29 2011Fletcher W J Sanchez Goni M F Peyron O and Dormoy

I Abrupt climate changes of the last deglaciation detected ina Western Mediterranean forest record Clim Past 6 245ndash264doi105194cp-6-245-2010 2010

Fletcher W J Debret M and Sanchez Goni M F Mid-Holoceneemergence of a low-frequency millennial oscillation in westernMediterranean climate implications for past dynamics of theNorth Atlantic atmospheric westerlies The Holocene 23 153ndash166 doi1011770959683612460783 2012

Franco P Jeftic L Malanotte Rizzoli P Michelato A and Or-lic M Descriptive model of the northern Adriatic OceanologicaActa 5 379ndash389 1982

Frignani M Langone L Pacelli M and Ravaioli M Input dis-tribution and accumulation of dolomite in sediments of the Mid-dle Adriatic Sea Rapp Comm Int Mer Medi 33 p 324 1992

Frignani M Langone l Ravaioli M Sorgente D Alvisi F andAlbertazzi S Fine-sediment mass balance in the western Adri-atic continental shelf over a century time scale Mar Geol 222ndash223 113ndash133 2005

Ginoux P Prospero J M Torres O and Chin M Long-termsimulation of global dust distribution with the GOCART modelcorrelation with North Atlantic Oscillation Environ ModellSoftw 19 113ndash128 2004

Giorgi F Climate change hot-spots Geophys Res Lett 33L08707 doi1010292006GL025734 2006

Giorgi F and Lionello P Climate change projections for theMediterranean region Global Planet Change 63 90ndash104 2008

Goudie A S and Middleton N J Saharan dust storms nature andconsequences Earth Sci Rev 56 179ndash204 2001

Grimm E C ldquoCONISS a FORTRAN 77 program for stratigraph-ically constrained cluster analysis by the method of incrementalsum of squaresrdquo Comput Geosci 13 13ndash35 1987

Guerzoni S and Chester R The Impact of Desert Dust Across theMediterranean Kluwer Academic Dordrecht 1996

Guerzoni S Chester R Dulac F Herut B Loye-Pilot M-DMeasures C Migon C Molinaroli E Moulin C RossiniP Saydam C Soudine A and Ziveri P The role of atmo-spheric deposition in the biogeochemistry of the MediterraneanSea Prog Oceanogr 44 147ndash190 1999

Guiot J Methodology of the last climatic cycle reconstruction inFrance from pollen data Palaeogeogr Palaeocl Palaeoecol 8049ndash69 1990

Heusser L E and Balsam W L Pollen distribution in the NEPacific ocean Quaternary Res 7 45ndash62 1977

Hicks S The use of annual arboreal pollen deposition values fordelimiting tree-lines in the landscape and exploring models ofpollen dispersal Rev Paleobotany and Palynology 117 SpecialIssue SI 1ndash29 doi101016S0034-6667(01)00074-4 2001

Hooghiemstra H Stalling H Agwu C O C and Dupont L MVegetational and climatic changes at the northern fringe of theSahara 250000ndash5000 years BP evidence from 4 marine pollenrecords located between Portugal and the Canary Islands RevPalaeobot Palynol 74 1ndash53 1992

Hooghiemstra H Leızine A-M Leroy S A G Dupont L andMarret F Late Quaternary palynology in marine sediments asynthesis of the understanding of pollen distribution patterns inthe NW African setting Quaternary Int 148 29ndash44 2006



Horvat I Glavac V and Ellenberg H Vegetation SudosteuropasGeobotanica Selecta 4 Stuttgart 1974

Hutson W The Agulhas current during the Late Pleistocene anal-ysis of modern faunal analogs Science 207 64ndash66 1980

IPCC Climate Change 2007 Synthesis Report Contribution ofWorking Groups I II and III to the Fourth Assessment Re-port of the Intergovernmental Panel on Climate Change GenevaSwitzerland 104 pp 2007

Jahns S and van der Bogaard C New palynological and tephros-tratigraphical investigations of two salt lagoons on the island ofMljet south Dalmatia Croatia Veget Hist Archaeobot 7 219ndash234 1998

Jalut G Esteban Amat A Bonnet L Gauquelin T andFontugne M Holocene climatic changes in the WesternMediterranean from south-east France to south-east SpainPalaeogeogr Palaeocl Palaeoecol 160 255ndash290 2000

Jalut G Dedoubat J J Fontugne M and Otto THolocene circum-Mediterranean vegetation changes Climateforcing and human impact Quaternary Int 200 4ndash18doi101016jquaint200803012 2009

Joannin S Brugiapaglia E de Beaulieu J L Bernado LMagny M Peyron O Goring S and Vanniere B Pollen-based reconstruction of Holocene vegetation and climate insouthern Italy the case of Lago di Trifoglietti Clim Past 81973ndash1996 2012httpwwwclim-pastnet819732012

Joannin S Vanniere B Galop D Peyron O Haas J NGilli A Chapron E Wirth S Anselmetti F Desmet Mand Magny M Climate and vegetation changes during theLateglacial and Early-Mid Holocene in Lake Ledro (southernAlps Italy) Clim Past 9 913ndash933 doi105194cp-9-913-2013 2013

Kotthoff U Muller U C Pross J Schmiedl G Lawson IT Van De Schootbrugge B and Schulz H Lateglacial andHolocene vegetation dynamics in the Aegean region An inte-grated view based on pollen data from marine and terrestrialarchives Holocene 18 1019ndash1032 2008a

Kotthoff U Pross J Muller U C Peyron O Schmiedl G andSchulz H Climate dynamics in the borderlands of the AegeanSea during formation of Sapropel S1 deduced from a marinepollen record Quaternary Sci Rev 27 832ndash845 2008b

Kotthoff U Koutsodendris A Pross J Schmiedl G Borne-mann A Kaul C Marino G Peyron O and Schiebel RImpact of late glacial cold events on the Northern Aegean regionreconstructed from marine and terrestrial proxy data J Quater-nary Sci 26 86ndash96 2011

Loye-Pilot M-D Martin J-M and Morelli J Influence of Sa-haran dust on the rainfall acidity and atmospheric input to theMediterranean Nature 321 427-428 1986

Magny M Peyron O Sadori L Ortu E Zanchetta GVanniere B and Tinner W Contrasting patterns of precipi-tation seasonality during the Holocene in the south- and north-central Mediterranean J Quaternary Sci 27 290ndash296 2012

Magny M Combourieu Nebout N de Beaulieu J L Bout-Roumazeilles V Colombaroli D Desprat S Francke AJoannin S Peyron O Revel M Sadori L Siani G Sicre MA Samartin S Simonneau A Tinner W Vanniere B Wag-ner B Zanchetta G Anselmetti F Brugiapaglia E ChapronE Debret M Desmet M Didier J Essallami L Galop D

Gilli A Haas J N Kallel N Millet L Stock A TuronJ L and Wirth S North-south palaeohydrological contrasts inthe central Mediterranean during the Holocene tentative synthe-sis and working hypotheses Clim Past Discuss 9 1901ndash1967doi105194cpd-9-1901-2013 2013

Magri D Late Quaternary vegetation history at Lagaccione nearLago di Bolsena (central Italy) Review Rev Palaeobot Palynol106 171ndash208 1999

Magri D and Sadori L Late Pleistocene and Holocene pollenstratigraphy at Lago di Vico (central Italy) Veget Hist Ar-chaeobot 8 247ndash260 1999

Mercuri A M Bandini Mazzanti M Florenzano A MontecchiM C and Rattighieri E Olea Juglans and Castanea The OJCgroup as pollen evidence of the development of human-inducedenvironments in the Italian peninsula Quaternary Int 303 24ndash42 doi101016jquaint201301005 2013

Milliman J D and Syvitski J P M Geomorphictectonic con-trol of sediment discharge to the ocean the importance of smallmountainous rivers J Geol 100 525ndash544 1992

Moulin C Lambert C E Dulac F and Dayan U Control ofatmospheric export of dust from North Africa by the North At-lantic Oscillation Nature 387 691ndash694 1997

New M Hulme M and Jones P D Representing twentieth cen-tury space-time climate variability Part 2 development of 1901ndash96 monthly grids of terrestrial surface climate J Climate 132217ndash2238 2000

Oldfield F Asioli A Accorsi C A Mercuri A M JugginsS Langone L Rolph T Trincardi F Wolff G Gibbs ZVigliotti L Frignani M van der Post K and Branch N Ahigh resolution late Holocene palaeoenvironmental record fromthe central Adriatic Sea Quatern Sci Rev 22 319ndash342 2003

Orange D Garcia-Garcia A Lorenson T Nittrouer C MilliganT Miserocchi S Langone L Corregiani A I and TrincardiF Shallow gas and flood deposition on the Po Delta Mar Geol222 159-177 2005

Ortu E David F and Peyron O Pollen-inferred palaeoclimatereconstruction in the Alps during the Lateglacial and the earlyHolocene how to estimate the effect of elevation and localparameters J Quat SCi 25 651ndash661 doi101002jqs13352010a

Ortu E Klotz S Brugiapaglia E Caramiello R and Sinis-calco C Elevation-induced variations of pollen assemblagesin the North-western Alps An analysis of their value as tem-perature indicators Comptes Rendus Biologie 333 11ndash12doi101016jcrvi201009006 2010b

Ozenda P Sur lesetages de vegetation dans les montagnes dubassin mediterraneen Documents de Cartographie Ecologique16 1ndash32 1975

Palinkas C M and Nittrouer C A Clinoform sedimentationalong the Apennines shelf Adriatic Sea Mar Geol 234 245ndash260 2007

Petschick R MacDiff v 425 (Free Geological Software) avail-able at httpwwwgeol-palunifrankfurtdeStaffHomepagesPetschickclassicsoftwarehtmlMacDiff(last access 2 July2012) Geologisch-Palaontologisches Institut Universitat Frank-furtMain 2001

Peyron O Guiot J Cheddadi R Tarasov P Reille M deBeaulieu J L Bottema S and Andrieu V Climatic re-construction in Europe for 18000 yr BP from pollen data



Quaternary Res 49 183ndash196 1998Peyron O Goring S Dormoy I Kotthoff U Pross J de

Beaulieu J L Drescher-Schneider R Vanniere B andMagny M Holocene seasonality changes in central Mediter-ranean reconstructed from Lake Accesa and Tenaghi Philipponpollen sequences Holocene 21 131ndash146 2011

Peyron O Magny M Goring S Joannin S De Beaulieu JL Brugiapaglia E Sadori L Garfi G Kouli K IoakimC and Combourieu-Nebout N Contrasting patterns of cli-matic changes during the Holocene in central Mediterranean area(Italy) reconstructed from pollen data Clim Past 9 1233ndash1252doi105194cp-9-1233-2013 2013

Piva A Asioli A Trincardi F Schneider R R and VigliottiL Late-Holocene climate variability in the Adriatic Sea (CentralMediterranean) Holocene 18 153ndash167 2008

Polunin O Flowers in Greece and the Balkan a field guide OxfordUniv Press 1980

Poulain P M Adriatic Sea surface circulation as derived fromdrifter data between 1990 and 1999 Mar Syst 29 3ndash32 2001

Prospero J M Arid regions as sources of minerals aerosols in themarine atmosphere Geo S Am S 186 71ndash86 1981

Pross J Kotthoff U Muller U C Peyron O Dormoy ISchmiedl G Kalaitzidis S and Smith A M Massive per-turbation in terrestrial ecosystems of the Eastern Mediterraneanregion associated with the 82 kyr BP climatic event Geology37 887ndash890 2009

Pye K Aeolian dust and dust deposits New-York London Aca-demic Press 334 pp 1987

Rasmussen S O Vinther B M Clausen H B and Ander-sen KK Early Holocene climate oscillations recorded in threeGreenland ice cores Quaternary Sci Rev 26 1907ndash1914 2007

Ravazzi C Late Quaternary history of spruce in southern EuropeRev Palaeobot Palynol 120 131ndash177 2002

Rea D K Leinen M and Jacenek T R Geologic approach tothe long-term history of atmospheric circulation Science 227721ndash725 1985

Reimer P J Baillie M G L Bard E Bayliss A Beck J WBertrand C J H Blackwell P G Buck C E Burr G SCutler K B Damon P E Edwards R L Fairbanks R GFriedrich M Guilderson T P Hogg A G Hughen K AKromer B McCormac F G Manning S W Ramsey C BReimer R W Remmele S Southon J R Stuiver M TalamoS Taylor F W van der Plicht J and Weyhenmeyer C E Int-Cal04 Terrestrial radiocarbon age calibration 26ndash0 ka BP Ra-diocarbon 46 1029ndash1058 2004

Revel M Ducassou E Grousset F Bernasconi S M MigeonS Revillon S Mascle J Murat A Zaragosi S and BoschD 100000 years of African monsoon variability recorded in sed-iments of the Nile margin Quaternary Sci Rev 29 1342ndash13622010

Rodriguez S Querol X Alastuey A Kallos G and KakaliagouO Saharan dust contributions to PM10 and TSP levels in South-ern and Eastern Spain Atmos Environ 35 2433ndash2447 2001

Sadori L and Narcisi B The postglacial record of environmentalhistory from Lago di Pergusa (Sicily) Holocene 11 655ndash6712001

Sadori L Jahns S and Peyron O Mid-Holocene vegetation his-tory of the Central Mediterranean Holocene 21 117ndash129 2011

Sadori L Ortu E Peyron O Zanchetta G Vanniere BDesmet M and Magny M The last 7 millennia of vegeta-tion and climate changes at Lago di Pergusa (central SicilyItaly) Clim Past Discuss 9 2059ndash2094 doi105194cpd-9-2059-2013 2013

Schmiedl G Kuhnt T Ehrmann W Emeis K Hamann YKotthoff U Dulski P and Pross J Climatic forcing of east-ern Mediterranean deep-water formation and benthic ecosystemsduring the past 22 000 years Quatern Sci Rev 29 3006ndash30202010

Siani G Paterne M Arnold M Bard E Metivier B TisneratN and Bassinot F Radiocarbon reservoir ages in the Mediter-ranean Sea and Black Sea coastal waters Radiocarbon 42 271ndash280 2000

Siani G Paterne M Michel E Sulpizio R Sbrana A ArnoldM and Haddad G Mediterranean Sea surface radiocarbonreservoir age changes since the Last Glacial Maximum Science294 1917ndash1920 doi101126science1063649 2001

Siani G Sulpizio R Paterne M and Sbrana A Tephrostratig-raphy study for the last 1800014C years in a deep-sea sedimentsequence for the South Adriatic Quaternary Sci Rev 23 2485ndash2500 2004

Siani G Paterne M and Colin C Late glacial to Holoceneplanktic foraminifera bioevents and climatic record in the SouthAdriatic Sea J Quaternary Sci 25 808ndash821 2010

Siani G Magny M and Paterne M Impact of the Holocene cli-mate variability on the Adriatic paleohydrology Clim Past 9499ndash515 doi105194cp-9-499-2013 2013

Sicre M-A Siani G Genty D Kallel N and Essallami LSeemingly divergent sea surface temperature proxy records in thecentral Mediterranean during the last deglaciation Clim Past 91375ndash1383 doi105194cp-9-1375-2013 2013

Sjogren P Knaap W V D Huusko A and Leeuwen J F VPollen productivity dispersal 22 and correction factors for ma-jor tree taxa in the Swiss Alps based on pollen-trap results RevPalaeobot Palynol 152 200ndash210 2008

Sorgente D Studio della sedimentazione attuale e recente nelmedio Adriatico attraverso lrsquouso di traccianti redioattivi PhDThesis University of Bologna 178 pp 1999

Ter Braak C J F and Juggins S Weighted averaging partial leastsquares regression (WA-PLS) An improved method for recon-structing environmental variables from species assemblages Hy-drobiologia 269270 485ndash502 1993

Tomadin L Sedimentary fluxed and different dispersion mecha-nism of the clay sediments in the Adriatic Basin Rend Fis AccLincei 9 161ndash174 2000

Tomadin L and Lenaz R Eolian dust over the Mediterranean andtheir contribution to the present sedimentation in Paleoclimatol-ogy and Paleometeorology modern and past patterns of globalatmospheric transport edited by Leinen M and Sarnthein MDordrecht Kluwer academic 282 267ndash281 1989

Torres-Padron M E Gelado-Caballero M D Collado-SanchezC Siruela-Matos V F Cardona-Castellano P J andHernandez-Brito J J Variability of dust inputs to the CANIGOzone Deep-Sea Res II 49 3455ndash3464 2002

Turon J-L Le palynoplancton dans lrsquoenvironnement actuel delrsquoAtlantique nord-oriental Evolution climatique et hydrologiquedepuis le dernier maximum glaciaire PhD thesis Bordeaux 1University France 1984



Tzedakis P C Seven ambiguities in the Mediterranean palaeoenvi-ronmental narrative Quaternary Sci Rev 26 2042ndash2066 2007

van der Knaap W O van Leeuwen J F N Finsinger W Go-bet E Pini R Schweizer A Valsecchi V and AmmannB Migration and population expansion of Abies Fagus Piceaand Quercus since 15000 years in and across the Alps basedon pollen-percentage threshold values Quaternary Sci Rev 24645ndash680 2005

Walker M J C Berkelhammer M Bjorck S Cwynar L CFisher D A Long A J Lowe J J Newnham R MRasmussen S O and Weiss H Formal subdivision of theHolocene SeriesEpoch a Discussion Paper by a Working Groupof INTIMATE (Integration of ice-core marine and terrestrialrecords) and the Subcommission on Quaternary Stratigraphy (In-ternational Commission on Stratigraphy) J Quaternary Sci 27640ndash659 2012

Walter W Harnickell E and Mueller-Dombois D Climate Dia-gramMaps Springer-Verlag 1975

Watts W A Allen J R M and Huntley B Vegetation historyand palaeoclimate of the Last Glacial period at Lago Grande diMonticchio Southern Italy Quaternary Sci Rev 15 133ndash1531996

Wheatcroft R A Nittrouer C A Miserocchi S and TrincardiF A comparison of recent floods and flood deposits of the Eeland Po Rivers in American Geophysical Union ndash Chapman con-ference Formation of sedimentary strata on continental marginsPuerto Rico 17ndash19 June 2001

Woodward F I Climate and plant distribution Cambridge UnivPress Cambridge England 1987

Wu H Guiot J Brewer S and Guo Z Climatic changesin Eurasia and Africa at the last glacial maximum andmid-Holocene reconstruction from pollen data using in-verse vegetation modelling Clim Dynam 29 211ndash229doi101007s00382-007-0231-3 2007

Zanchettin D Traverso P and Tomasino M Po River dis-charges a preliminary analysis of a 200-year tile series ClimChange 89 411ndash433 2008

Zhornyak L V Zanchetta G Drysdale R N Hellstrom JC Isola I Regattieri E Piccini L Baneschi I and Cou-choud I Stratigraphic evidence for a ldquopluvial phaserdquo betweenca 8200ndash7100 ka from Renella cave (Central Italy) QuaternarySci Rev 30 409ndash417 2011



1 Introduction

The Mediterranean is highly sensitive to climate change andMediterranean ecosystems already widely affected by an-thropogenic pressures are likely to be strongly impacted byfuture warming and drought stresses Precipitation changesare of particular concern for the region Increasing dry-ness predicted by models for the Mediterranean will in-duce large water resource deficits and might threaten the re-gionrsquos habitability (Giorgi 2006 IPCC 2007 Giorgi andLionello 2008) While modern Mediterranean precipitationis controlled by the influences of both subtropical and mid-latitudinal climatic belts resulting in dry summers and rainywinters climate in the Mediterranean has changed during theHolocene providing an opportunity to study ecosystem re-sponse under a range of climate conditions Thus reconstruc-tions of past vegetation and climate in the different basinsof the Mediterranean will produce information allowing usto draw a general framework to understand regional changeand the drivers of environmental and climate change acrossthe whole basin

The southern Adriatic with its central location in theMediterranean Basin provides an excellent opportunity torecord high-resolution vegetation changes in Italy and theBalkans in response to Holocene climate variability In factcentral Mediterranean ecosystems are highly sensitive to cli-mate changes and help to understand the climate connectionsand conflicting influences between northwestern and south-eastern atmospheric systems and their relative roles in medi-ating precipitation and runoff during the Holocene

Holocene vegetation change in the central Mediterraneanhas been widely investigated using continental and marinerecords (eg Watts et al 1996 Combourieu-Nebout et al1998 Jahns and van den Bogaard 1998 Magri 1999 Ma-gri and Sadori 1999 Allen et al 2002 Oldfield et al 2003Drescher-Schneider et al 2007 Piva et al 2008 Kotthoff etal 2008a b 2011 Di Donato et al 2008 Di Rita and Ma-gri 2009 Joannin et al 2012 2013 Di Rita et al 2013) Al-though these records vary in temporal and taxonomic resolu-tion they show a general pattern of forest expansion linked towarming climates in the Holocene as well as the influence ofmillennial-scale climate variability on continental paleoenvi-ronments Most records from the central Mediterranean di-vide the Holocene into three phases similar to those in otherregions of the Mediterranean (eg Jalut et al 2000 2009Finne et al 2011) with boundaries that have been recentlyrefined by Walker et al (2012) (i) an early Holocene humidperiod from 11 700 to 8200 cal yr BP (calendar years beforepresent) (ii) a mid-Holocene transitional period from 8200to 4200 cal yr BP and (iii) a late-Holocene period of arid-ification following 4200 cal yr BP Climate reconstructionsfrom Lake Accesa (northcentral Italy) distinguish two dis-tinct intervals based on precipitation and emphasize the roleof the seasonality during the Holocene (Peyron et al 2011)Given the scarcity of reliable high-resolution pollen records

in the central Mediterranean and the peculiarities of this re-gion new evidence is needed to better understand the climatechanges and the climate processes in this area

The new pollen record from the marine MD 90-917 coreprovides new insights into vegetation and climate changesfor the central Mediterranean This study part of the Frenchmulti-proxy project LAMA provides a new understanding ofthe links between hydrological variability climate and vege-tation in the central Mediterranean by bringing together con-tinental and marine data to produce an integrated interpre-tation of the paleoenvironmental interpretation of Holoceneclimate changes This paper investigates the impacts oflong-term and millennial-scale climate variability during theHolocene through temperature and precipitation stresses onItalian and Balkan vegetation By examining the clay fractionratios from the same core we will be able to discuss patternsof local and regional runoff linked to precipitation and sea-sonality changes

2 Lithology and age model

Core MD 90-917 (41 N 1737prime E) was collected in the Adri-atic Sea by the RVMarion Dufresneat a depth of 1010 m(Fig 1) The core is described more fully in Combourieu-Nebout et al (1998) and Siani et al (2004 2010) Sedimentsare composed of 21 m of clay interrupted by two black lay-ers corresponding to the deposition of the S1 sapropel in twolevels S1a and S1b respectively at 249ndash255 cm and 229ndash239 cm downcore (Siani et al 2004) The pollen record pre-sented in this paper focuses on the upper 33 m of the core

The age model of the core was previously given in Sianiet al (2010) and uses tephra layers and 21 AMS (accelera-tor mass spectrometry)14C dates performed on monospecificplanktonic foraminifera The conventional radiocarbon ageshave been subsequently converted into calendar ages basedon INTCAL04 (Reimer et al 2004) using the14C calibrationsoftware CALIB 6 The calibration integrates a marine14Creservoir age correctionR(t) at 390plusmn85 yr according to Sianiet al (2000 2001) For this paper only 13 AMS14C are usedand provide a reliable age model for the last 13 000 cal yr BP(Table 1)

3 Present-day environmental settings

31 Climate and atmospheric circulation pattern

Atmospheric circulation in the Adriatic is controlled by twodominant climatic systems In the winter the Eurasian Highextends to the southwest The Eurasian High retreats north-ward in the summer and is replaced by the northeastwardexpansion of the Azores High This pattern results in a typ-ically Mediterranean climate (Fig 1) winters are generallystormy and coolcold while summers are warmhot and drySpring climate in the region is generally long characterized



0

10

20

30

40

50

0

40

80

120

160

200


BoraMaestral

Jugo

OstroSirocco

Libeccio

ROMA

TUNIS

MD 04-2797

VENICE

4

Trifoglietti

Ledro

Pergusa

Accesa

Maliq

Monticchio



MD 90-917

5

1

62

3

0

10

20

30

40

50

0

40

80

120

160

200



BARI 159 586 49m

0

10

20

30

40

50

0

40

80

120

160

200



0

10

20

30

40

50

0

40

80

120

160

200


SPLIT 158 893 128m

0

10

20

30

40

50

0

40

80

120

160

200



0

10

20

30

40

50

0

40

80

120

160

200


DURRES 159 962 0m


1

Preola

4

2 5

3 6


Otranto strait

Po
















4 Methods

41 Pollen record







(b)









uarr

seasonalP mm

































Steppe Cold and dry











51 Pollen inputs





MTCO degC

MTWA degC

PANN mmTANN degC

Psum mm

Pwin mm


8200

yr e

vent

Youn

ger D

ryas

-8

-4

0

4

8

2000 4000 6000 8000 10000 12000

10

15

20

25

30

2000 4000 6000 8000 10000 12000

5

10

15

20

2000 4000 6000 8000 10000 12000200

400

600

800

2000 4000 6000 8000 10000 12000

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

0

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

2000 4000 6000 8000 10000 12000Age (cal yr BP)

PB

OP

B

2000 4000 6000 8000 10000 12000Age (cal yr BP)

4200

yr

even

t

10

12

14

18

16

10

12

14

18

16

Youn

ger D

ryas

PB

OP


8200

yr e

vent



















0

10

20

30

40

50

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

0

10

20

30

400

500

600

700

800

900

1000

300

100

200100

200

100

200

300

100

200

60

80

70

0

10

20

30

40

50

60

80

70

0

10

20

0

10

20

30

0

10

20

30

40

0

10

Alti

tudi

nal

taxa

()

Oth

er te

mpe

rate

Cor

ylus

()

Tem

pera

te fo

rest

d

ecid

uous

Q

uerc

us (

)Med

iterr

anea

nta

xa (

)S

emi-d

eser

tta

xa (

)P

sum

(mm

)

Her

bace

ous

taxa

()

Psp

r (m

m)

Pau

t (m

m)

PAN

N (m

m)

Pw

in (m

m)

Mea

n te

mpe

ratu

re (deg

C)MTWA

MTCO

TANN

Age (cal yr BP)

Age (cal yr BP)

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

917-V

917-X

II

917-X

I

917-X

917-I

X

917-V

IIIb

917-V

I

917-I

V

917-I

II

917-I

917-I

I

917-V

IIIa

917-V

II

Pol

len

zona

tion

Present-day value




-40

-34

0

20

40

60

80

2000 4000 6000 8000 10000 12000

0

60

40

20

2000 4000 6000 8000 10000 12000

Age (cal yr BP)


Age (cal yr BP)

Semi desert ()

Temperate forest ()

-36

-38

-42

82

00 y

r eve

nt

PB

O

Youn

ger

Dry

as

PB

18

16

14

12

10









- 250

- 150

- 050

20

40

60

80

100

2000 4000 6000 8000 10000 12000 14000

050

150

250

Trees

18δ O permil

13Cδ C permil

Age (cal yr BP)

18

810121416
















1000 3000 1100050002000 4000 7000 80000 100009000 120006000

-4

-6

-8

-10

-12

2

0

-2

-4

-6MTW

A an

omal

ies

MTC

O a

nom

alie

s

MD 90-917

SW Europe

SW Europe

SE Europe

SE Europe

MD 90-917

1000 3000 1100050002000 4000 7000 80000 100009000 120006000

1

0

-1

-2

-3

2

1

0

-1

-2

0

-1

-2

-3

-4

3

2

1

0

Age (cal yr BP)

Age (cal yr BP)

MTW

A anomalies

MTC

O anom

aliesM

TWA anom

aliesM

TCO

anomalies






Combourieu-

YD YD

SL 152



200 40 60

Quercus ()

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age

(cal

yr B

P)

ACP2

ACP3

ACP4

ACP5

ACP6

ACP7

ACP8ACP9

YD

NAC1

NAC2

NAC3

NAC4

NAC5

NAC6

NAC7

NAC8

NAYD

MD 90-917


(2001)


cold events

YD

West-CentralEurope

YD

AdC1

AdC2

AdC4

AdC6

AdC7

AdC8

AdC3

AdC5

MD 01-2797

Desprat et al(2013)

MD 99-2043


Magny et al(2012)

ODP 976

Nebout et al(2009)

60


PBO

BO

8200

9200

4200

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age (cal yr BP)









2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

100

200Psum (mm)

S1aS1b

Age (cal yr BP)


+

-

2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

Age (cal yr BP)

-6


IK

Po

Riv

er

disc

harg

e

+

-

loca

l riv

ers

dis

char

ge

-

+

salin

e

+

- Sum

mer

Pre

cipi

tatio

n

-52

-49

-46

-43

-40

-37

15

20

1

2

3


- Pre

cipi

tatio

n

+

- Pre

cipi

tatio

n

3

5

7

9

Pwin (mm)100

200 +

- Win

ter

Pre

cipi

tatio

n

SK


PBOBO82 92 42

M

D

90

917





8 Conclusions











References




































































































































0

10

20

30

40

50

0

40

80

120

160

200


BoraMaestral

Jugo

OstroSirocco

Libeccio

ROMA

TUNIS

MD 04-2797

VENICE

4

Trifoglietti

Ledro

Pergusa

Accesa

Maliq

Monticchio



MD 90-917

5

1

62

3

0

10

20

30

40

50

0

40

80

120

160

200



BARI 159 586 49m

0

10

20

30

40

50

0

40

80

120

160

200



0

10

20

30

40

50

0

40

80

120

160

200


SPLIT 158 893 128m

0

10

20

30

40

50

0

40

80

120

160

200



0

10

20

30

40

50

0

40

80

120

160

200


DURRES 159 962 0m


1

Preola

4

2 5

3 6


Otranto strait

Po
















4 Methods

41 Pollen record







(b)









uarr

seasonalP mm

































Steppe Cold and dry











51 Pollen inputs





MTCO degC

MTWA degC

PANN mmTANN degC

Psum mm

Pwin mm


8200

yr e

vent

Youn

ger D

ryas

-8

-4

0

4

8

2000 4000 6000 8000 10000 12000

10

15

20

25

30

2000 4000 6000 8000 10000 12000

5

10

15

20

2000 4000 6000 8000 10000 12000200

400

600

800

2000 4000 6000 8000 10000 12000

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

0

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

2000 4000 6000 8000 10000 12000Age (cal yr BP)

PB

OP

B

2000 4000 6000 8000 10000 12000Age (cal yr BP)

4200

yr

even

t

10

12

14

18

16

10

12

14

18

16

Youn

ger D

ryas

PB

OP


8200

yr e

vent



















0

10

20

30

40

50

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

0

10

20

30

400

500

600

700

800

900

1000

300

100

200100

200

100

200

300

100

200

60

80

70

0

10

20

30

40

50

60

80

70

0

10

20

0

10

20

30

0

10

20

30

40

0

10

Alti

tudi

nal

taxa

()

Oth

er te

mpe

rate

Cor

ylus

()

Tem

pera

te fo

rest

d

ecid

uous

Q

uerc

us (

)Med

iterr

anea

nta

xa (

)S

emi-d

eser

tta

xa (

)P

sum

(mm

)

Her

bace

ous

taxa

()

Psp

r (m

m)

Pau

t (m

m)

PAN

N (m

m)

Pw

in (m

m)

Mea

n te

mpe

ratu

re (deg

C)MTWA

MTCO

TANN

Age (cal yr BP)

Age (cal yr BP)

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

917-V

917-X

II

917-X

I

917-X

917-I

X

917-V

IIIb

917-V

I

917-I

V

917-I

II

917-I

917-I

I

917-V

IIIa

917-V

II

Pol

len

zona

tion

Present-day value




-40

-34

0

20

40

60

80

2000 4000 6000 8000 10000 12000

0

60

40

20

2000 4000 6000 8000 10000 12000

Age (cal yr BP)


Age (cal yr BP)

Semi desert ()

Temperate forest ()

-36

-38

-42

82

00 y

r eve

nt

PB

O

Youn

ger

Dry

as

PB

18

16

14

12

10









- 250

- 150

- 050

20

40

60

80

100

2000 4000 6000 8000 10000 12000 14000

050

150

250

Trees

18δ O permil

13Cδ C permil

Age (cal yr BP)

18

810121416
















1000 3000 1100050002000 4000 7000 80000 100009000 120006000

-4

-6

-8

-10

-12

2

0

-2

-4

-6MTW

A an

omal

ies

MTC

O a

nom

alie

s

MD 90-917

SW Europe

SW Europe

SE Europe

SE Europe

MD 90-917

1000 3000 1100050002000 4000 7000 80000 100009000 120006000

1

0

-1

-2

-3

2

1

0

-1

-2

0

-1

-2

-3

-4

3

2

1

0

Age (cal yr BP)

Age (cal yr BP)

MTW

A anomalies

MTC

O anom

aliesM

TWA anom

aliesM

TCO

anomalies






Combourieu-

YD YD

SL 152



200 40 60

Quercus ()

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age

(cal

yr B

P)

ACP2

ACP3

ACP4

ACP5

ACP6

ACP7

ACP8ACP9

YD

NAC1

NAC2

NAC3

NAC4

NAC5

NAC6

NAC7

NAC8

NAYD

MD 90-917


(2001)


cold events

YD

West-CentralEurope

YD

AdC1

AdC2

AdC4

AdC6

AdC7

AdC8

AdC3

AdC5

MD 01-2797

Desprat et al(2013)

MD 99-2043


Magny et al(2012)

ODP 976

Nebout et al(2009)

60


PBO

BO

8200

9200

4200

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age (cal yr BP)









2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

100

200Psum (mm)

S1aS1b

Age (cal yr BP)


+

-

2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

Age (cal yr BP)

-6


IK

Po

Riv

er

disc

harg

e

+

-

loca

l riv

ers

dis

char

ge

-

+

salin

e

+

- Sum

mer

Pre

cipi

tatio

n

-52

-49

-46

-43

-40

-37

15

20

1

2

3


- Pre

cipi

tatio

n

+

- Pre

cipi

tatio

n

3

5

7

9

Pwin (mm)100

200 +

- Win

ter

Pre

cipi

tatio

n

SK


PBOBO82 92 42

M

D

90

917





8 Conclusions











References












































































































































4 Methods

41 Pollen record







(b)









uarr

seasonalP mm

































Steppe Cold and dry











51 Pollen inputs





MTCO degC

MTWA degC

PANN mmTANN degC

Psum mm

Pwin mm


8200

yr e

vent

Youn

ger D

ryas

-8

-4

0

4

8

2000 4000 6000 8000 10000 12000

10

15

20

25

30

2000 4000 6000 8000 10000 12000

5

10

15

20

2000 4000 6000 8000 10000 12000200

400

600

800

2000 4000 6000 8000 10000 12000

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

0

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

2000 4000 6000 8000 10000 12000Age (cal yr BP)

PB

OP

B

2000 4000 6000 8000 10000 12000Age (cal yr BP)

4200

yr

even

t

10

12

14

18

16

10

12

14

18

16

Youn

ger D

ryas

PB

OP


8200

yr e

vent



















0

10

20

30

40

50

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

0

10

20

30

400

500

600

700

800

900

1000

300

100

200100

200

100

200

300

100

200

60

80

70

0

10

20

30

40

50

60

80

70

0

10

20

0

10

20

30

0

10

20

30

40

0

10

Alti

tudi

nal

taxa

()

Oth

er te

mpe

rate

Cor

ylus

()

Tem

pera

te fo

rest

d

ecid

uous

Q

uerc

us (

)Med

iterr

anea

nta

xa (

)S

emi-d

eser

tta

xa (

)P

sum

(mm

)

Her

bace

ous

taxa

()

Psp

r (m

m)

Pau

t (m

m)

PAN

N (m

m)

Pw

in (m

m)

Mea

n te

mpe

ratu

re (deg

C)MTWA

MTCO

TANN

Age (cal yr BP)

Age (cal yr BP)

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

917-V

917-X

II

917-X

I

917-X

917-I

X

917-V

IIIb

917-V

I

917-I

V

917-I

II

917-I

917-I

I

917-V

IIIa

917-V

II

Pol

len

zona

tion

Present-day value




-40

-34

0

20

40

60

80

2000 4000 6000 8000 10000 12000

0

60

40

20

2000 4000 6000 8000 10000 12000

Age (cal yr BP)


Age (cal yr BP)

Semi desert ()

Temperate forest ()

-36

-38

-42

82

00 y

r eve

nt

PB

O

Youn

ger

Dry

as

PB

18

16

14

12

10









- 250

- 150

- 050

20

40

60

80

100

2000 4000 6000 8000 10000 12000 14000

050

150

250

Trees

18δ O permil

13Cδ C permil

Age (cal yr BP)

18

810121416
















1000 3000 1100050002000 4000 7000 80000 100009000 120006000

-4

-6

-8

-10

-12

2

0

-2

-4

-6MTW

A an

omal

ies

MTC

O a

nom

alie

s

MD 90-917

SW Europe

SW Europe

SE Europe

SE Europe

MD 90-917

1000 3000 1100050002000 4000 7000 80000 100009000 120006000

1

0

-1

-2

-3

2

1

0

-1

-2

0

-1

-2

-3

-4

3

2

1

0

Age (cal yr BP)

Age (cal yr BP)

MTW

A anomalies

MTC

O anom

aliesM

TWA anom

aliesM

TCO

anomalies






Combourieu-

YD YD

SL 152



200 40 60

Quercus ()

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age

(cal

yr B

P)

ACP2

ACP3

ACP4

ACP5

ACP6

ACP7

ACP8ACP9

YD

NAC1

NAC2

NAC3

NAC4

NAC5

NAC6

NAC7

NAC8

NAYD

MD 90-917


(2001)


cold events

YD

West-CentralEurope

YD

AdC1

AdC2

AdC4

AdC6

AdC7

AdC8

AdC3

AdC5

MD 01-2797

Desprat et al(2013)

MD 99-2043


Magny et al(2012)

ODP 976

Nebout et al(2009)

60


PBO

BO

8200

9200

4200

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age (cal yr BP)









2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

100

200Psum (mm)

S1aS1b

Age (cal yr BP)


+

-

2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

Age (cal yr BP)

-6


IK

Po

Riv

er

disc

harg

e

+

-

loca

l riv

ers

dis

char

ge

-

+

salin

e

+

- Sum

mer

Pre

cipi

tatio

n

-52

-49

-46

-43

-40

-37

15

20

1

2

3


- Pre

cipi

tatio

n

+

- Pre

cipi

tatio

n

3

5

7

9

Pwin (mm)100

200 +

- Win

ter

Pre

cipi

tatio

n

SK


PBOBO82 92 42

M

D

90

917





8 Conclusions











References




































































































































(b)









uarr

seasonalP mm

































Steppe Cold and dry











51 Pollen inputs





MTCO degC

MTWA degC

PANN mmTANN degC

Psum mm

Pwin mm


8200

yr e

vent

Youn

ger D

ryas

-8

-4

0

4

8

2000 4000 6000 8000 10000 12000

10

15

20

25

30

2000 4000 6000 8000 10000 12000

5

10

15

20

2000 4000 6000 8000 10000 12000200

400

600

800

2000 4000 6000 8000 10000 12000

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

0

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

2000 4000 6000 8000 10000 12000Age (cal yr BP)

PB

OP

B

2000 4000 6000 8000 10000 12000Age (cal yr BP)

4200

yr

even

t

10

12

14

18

16

10

12

14

18

16

Youn

ger D

ryas

PB

OP


8200

yr e

vent



















0

10

20

30

40

50

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

0

10

20

30

400

500

600

700

800

900

1000

300

100

200100

200

100

200

300

100

200

60

80

70

0

10

20

30

40

50

60

80

70

0

10

20

0

10

20

30

0

10

20

30

40

0

10

Alti

tudi

nal

taxa

()

Oth

er te

mpe

rate

Cor

ylus

()

Tem

pera

te fo

rest

d

ecid

uous

Q

uerc

us (

)Med

iterr

anea

nta

xa (

)S

emi-d

eser

tta

xa (

)P

sum

(mm

)

Her

bace

ous

taxa

()

Psp

r (m

m)

Pau

t (m

m)

PAN

N (m

m)

Pw

in (m

m)

Mea

n te

mpe

ratu

re (deg

C)MTWA

MTCO

TANN

Age (cal yr BP)

Age (cal yr BP)

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

917-V

917-X

II

917-X

I

917-X

917-I

X

917-V

IIIb

917-V

I

917-I

V

917-I

II

917-I

917-I

I

917-V

IIIa

917-V

II

Pol

len

zona

tion

Present-day value




-40

-34

0

20

40

60

80

2000 4000 6000 8000 10000 12000

0

60

40

20

2000 4000 6000 8000 10000 12000

Age (cal yr BP)


Age (cal yr BP)

Semi desert ()

Temperate forest ()

-36

-38

-42

82

00 y

r eve

nt

PB

O

Youn

ger

Dry

as

PB

18

16

14

12

10









- 250

- 150

- 050

20

40

60

80

100

2000 4000 6000 8000 10000 12000 14000

050

150

250

Trees

18δ O permil

13Cδ C permil

Age (cal yr BP)

18

810121416
















1000 3000 1100050002000 4000 7000 80000 100009000 120006000

-4

-6

-8

-10

-12

2

0

-2

-4

-6MTW

A an

omal

ies

MTC

O a

nom

alie

s

MD 90-917

SW Europe

SW Europe

SE Europe

SE Europe

MD 90-917

1000 3000 1100050002000 4000 7000 80000 100009000 120006000

1

0

-1

-2

-3

2

1

0

-1

-2

0

-1

-2

-3

-4

3

2

1

0

Age (cal yr BP)

Age (cal yr BP)

MTW

A anomalies

MTC

O anom

aliesM

TWA anom

aliesM

TCO

anomalies






Combourieu-

YD YD

SL 152



200 40 60

Quercus ()

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age

(cal

yr B

P)

ACP2

ACP3

ACP4

ACP5

ACP6

ACP7

ACP8ACP9

YD

NAC1

NAC2

NAC3

NAC4

NAC5

NAC6

NAC7

NAC8

NAYD

MD 90-917


(2001)


cold events

YD

West-CentralEurope

YD

AdC1

AdC2

AdC4

AdC6

AdC7

AdC8

AdC3

AdC5

MD 01-2797

Desprat et al(2013)

MD 99-2043


Magny et al(2012)

ODP 976

Nebout et al(2009)

60


PBO

BO

8200

9200

4200

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age (cal yr BP)









2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

100

200Psum (mm)

S1aS1b

Age (cal yr BP)


+

-

2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

Age (cal yr BP)

-6


IK

Po

Riv

er

disc

harg

e

+

-

loca

l riv

ers

dis

char

ge

-

+

salin

e

+

- Sum

mer

Pre

cipi

tatio

n

-52

-49

-46

-43

-40

-37

15

20

1

2

3


- Pre

cipi

tatio

n

+

- Pre

cipi

tatio

n

3

5

7

9

Pwin (mm)100

200 +

- Win

ter

Pre

cipi

tatio

n

SK


PBOBO82 92 42

M

D

90

917





8 Conclusions











References









































































































































uarr

seasonalP mm

































Steppe Cold and dry











51 Pollen inputs





MTCO degC

MTWA degC

PANN mmTANN degC

Psum mm

Pwin mm


8200

yr e

vent

Youn

ger D

ryas

-8

-4

0

4

8

2000 4000 6000 8000 10000 12000

10

15

20

25

30

2000 4000 6000 8000 10000 12000

5

10

15

20

2000 4000 6000 8000 10000 12000200

400

600

800

2000 4000 6000 8000 10000 12000

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

0

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

2000 4000 6000 8000 10000 12000Age (cal yr BP)

PB

OP

B

2000 4000 6000 8000 10000 12000Age (cal yr BP)

4200

yr

even

t

10

12

14

18

16

10

12

14

18

16

Youn

ger D

ryas

PB

OP


8200

yr e

vent



















0

10

20

30

40

50

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

0

10

20

30

400

500

600

700

800

900

1000

300

100

200100

200

100

200

300

100

200

60

80

70

0

10

20

30

40

50

60

80

70

0

10

20

0

10

20

30

0

10

20

30

40

0

10

Alti

tudi

nal

taxa

()

Oth

er te

mpe

rate

Cor

ylus

()

Tem

pera

te fo

rest

d

ecid

uous

Q

uerc

us (

)Med

iterr

anea

nta

xa (

)S

emi-d

eser

tta

xa (

)P

sum

(mm

)

Her

bace

ous

taxa

()

Psp

r (m

m)

Pau

t (m

m)

PAN

N (m

m)

Pw

in (m

m)

Mea

n te

mpe

ratu

re (deg

C)MTWA

MTCO

TANN

Age (cal yr BP)

Age (cal yr BP)

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

917-V

917-X

II

917-X

I

917-X

917-I

X

917-V

IIIb

917-V

I

917-I

V

917-I

II

917-I

917-I

I

917-V

IIIa

917-V

II

Pol

len

zona

tion

Present-day value




-40

-34

0

20

40

60

80

2000 4000 6000 8000 10000 12000

0

60

40

20

2000 4000 6000 8000 10000 12000

Age (cal yr BP)


Age (cal yr BP)

Semi desert ()

Temperate forest ()

-36

-38

-42

82

00 y

r eve

nt

PB

O

Youn

ger

Dry

as

PB

18

16

14

12

10









- 250

- 150

- 050

20

40

60

80

100

2000 4000 6000 8000 10000 12000 14000

050

150

250

Trees

18δ O permil

13Cδ C permil

Age (cal yr BP)

18

810121416
















1000 3000 1100050002000 4000 7000 80000 100009000 120006000

-4

-6

-8

-10

-12

2

0

-2

-4

-6MTW

A an

omal

ies

MTC

O a

nom

alie

s

MD 90-917

SW Europe

SW Europe

SE Europe

SE Europe

MD 90-917

1000 3000 1100050002000 4000 7000 80000 100009000 120006000

1

0

-1

-2

-3

2

1

0

-1

-2

0

-1

-2

-3

-4

3

2

1

0

Age (cal yr BP)

Age (cal yr BP)

MTW

A anomalies

MTC

O anom

aliesM

TWA anom

aliesM

TCO

anomalies






Combourieu-

YD YD

SL 152



200 40 60

Quercus ()

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age

(cal

yr B

P)

ACP2

ACP3

ACP4

ACP5

ACP6

ACP7

ACP8ACP9

YD

NAC1

NAC2

NAC3

NAC4

NAC5

NAC6

NAC7

NAC8

NAYD

MD 90-917


(2001)


cold events

YD

West-CentralEurope

YD

AdC1

AdC2

AdC4

AdC6

AdC7

AdC8

AdC3

AdC5

MD 01-2797

Desprat et al(2013)

MD 99-2043


Magny et al(2012)

ODP 976

Nebout et al(2009)

60


PBO

BO

8200

9200

4200

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age (cal yr BP)









2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

100

200Psum (mm)

S1aS1b

Age (cal yr BP)


+

-

2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

Age (cal yr BP)

-6


IK

Po

Riv

er

disc

harg

e

+

-

loca

l riv

ers

dis

char

ge

-

+

salin

e

+

- Sum

mer

Pre

cipi

tatio

n

-52

-49

-46

-43

-40

-37

15

20

1

2

3


- Pre

cipi

tatio

n

+

- Pre

cipi

tatio

n

3

5

7

9

Pwin (mm)100

200 +

- Win

ter

Pre

cipi

tatio

n

SK


PBOBO82 92 42

M

D

90

917





8 Conclusions











References












































































































































51 Pollen inputs





MTCO degC

MTWA degC

PANN mmTANN degC

Psum mm

Pwin mm


8200

yr e

vent

Youn

ger D

ryas

-8

-4

0

4

8

2000 4000 6000 8000 10000 12000

10

15

20

25

30

2000 4000 6000 8000 10000 12000

5

10

15

20

2000 4000 6000 8000 10000 12000200

400

600

800

2000 4000 6000 8000 10000 12000

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

0

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

2000 4000 6000 8000 10000 12000Age (cal yr BP)

PB

OP

B

2000 4000 6000 8000 10000 12000Age (cal yr BP)

4200

yr

even

t

10

12

14

18

16

10

12

14

18

16

Youn

ger D

ryas

PB

OP


8200

yr e

vent



















0

10

20

30

40

50

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

0

10

20

30

400

500

600

700

800

900

1000

300

100

200100

200

100

200

300

100

200

60

80

70

0

10

20

30

40

50

60

80

70

0

10

20

0

10

20

30

0

10

20

30

40

0

10

Alti

tudi

nal

taxa

()

Oth

er te

mpe

rate

Cor

ylus

()

Tem

pera

te fo

rest

d

ecid

uous

Q

uerc

us (

)Med

iterr

anea

nta

xa (

)S

emi-d

eser

tta

xa (

)P

sum

(mm

)

Her

bace

ous

taxa

()

Psp

r (m

m)

Pau

t (m

m)

PAN

N (m

m)

Pw

in (m

m)

Mea

n te

mpe

ratu

re (deg

C)MTWA

MTCO

TANN

Age (cal yr BP)

Age (cal yr BP)

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

917-V

917-X

II

917-X

I

917-X

917-I

X

917-V

IIIb

917-V

I

917-I

V

917-I

II

917-I

917-I

I

917-V

IIIa

917-V

II

Pol

len

zona

tion

Present-day value




-40

-34

0

20

40

60

80

2000 4000 6000 8000 10000 12000

0

60

40

20

2000 4000 6000 8000 10000 12000

Age (cal yr BP)


Age (cal yr BP)

Semi desert ()

Temperate forest ()

-36

-38

-42

82

00 y

r eve

nt

PB

O

Youn

ger

Dry

as

PB

18

16

14

12

10









- 250

- 150

- 050

20

40

60

80

100

2000 4000 6000 8000 10000 12000 14000

050

150

250

Trees

18δ O permil

13Cδ C permil

Age (cal yr BP)

18

810121416
















1000 3000 1100050002000 4000 7000 80000 100009000 120006000

-4

-6

-8

-10

-12

2

0

-2

-4

-6MTW

A an

omal

ies

MTC

O a

nom

alie

s

MD 90-917

SW Europe

SW Europe

SE Europe

SE Europe

MD 90-917

1000 3000 1100050002000 4000 7000 80000 100009000 120006000

1

0

-1

-2

-3

2

1

0

-1

-2

0

-1

-2

-3

-4

3

2

1

0

Age (cal yr BP)

Age (cal yr BP)

MTW

A anomalies

MTC

O anom

aliesM

TWA anom

aliesM

TCO

anomalies






Combourieu-

YD YD

SL 152



200 40 60

Quercus ()

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age

(cal

yr B

P)

ACP2

ACP3

ACP4

ACP5

ACP6

ACP7

ACP8ACP9

YD

NAC1

NAC2

NAC3

NAC4

NAC5

NAC6

NAC7

NAC8

NAYD

MD 90-917


(2001)


cold events

YD

West-CentralEurope

YD

AdC1

AdC2

AdC4

AdC6

AdC7

AdC8

AdC3

AdC5

MD 01-2797

Desprat et al(2013)

MD 99-2043


Magny et al(2012)

ODP 976

Nebout et al(2009)

60


PBO

BO

8200

9200

4200

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age (cal yr BP)









2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

100

200Psum (mm)

S1aS1b

Age (cal yr BP)


+

-

2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

Age (cal yr BP)

-6


IK

Po

Riv

er

disc

harg

e

+

-

loca

l riv

ers

dis

char

ge

-

+

salin

e

+

- Sum

mer

Pre

cipi

tatio

n

-52

-49

-46

-43

-40

-37

15

20

1

2

3


- Pre

cipi

tatio

n

+

- Pre

cipi

tatio

n

3

5

7

9

Pwin (mm)100

200 +

- Win

ter

Pre

cipi

tatio

n

SK


PBOBO82 92 42

M

D

90

917





8 Conclusions











References




































































































































MTCO degC

MTWA degC

PANN mmTANN degC

Psum mm

Pwin mm


8200

yr e

vent

Youn

ger D

ryas

-8

-4

0

4

8

2000 4000 6000 8000 10000 12000

10

15

20

25

30

2000 4000 6000 8000 10000 12000

5

10

15

20

2000 4000 6000 8000 10000 12000200

400

600

800

2000 4000 6000 8000 10000 12000

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

0

50

100

150

200

250

300

2000 4000 6000 8000 10000 12000

2000 4000 6000 8000 10000 12000Age (cal yr BP)

PB

OP

B

2000 4000 6000 8000 10000 12000Age (cal yr BP)

4200

yr

even

t

10

12

14

18

16

10

12

14

18

16

Youn

ger D

ryas

PB

OP


8200

yr e

vent



















0

10

20

30

40

50

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

0

10

20

30

400

500

600

700

800

900

1000

300

100

200100

200

100

200

300

100

200

60

80

70

0

10

20

30

40

50

60

80

70

0

10

20

0

10

20

30

0

10

20

30

40

0

10

Alti

tudi

nal

taxa

()

Oth

er te

mpe

rate

Cor

ylus

()

Tem

pera

te fo

rest

d

ecid

uous

Q

uerc

us (

)Med

iterr

anea

nta

xa (

)S

emi-d

eser

tta

xa (

)P

sum

(mm

)

Her

bace

ous

taxa

()

Psp

r (m

m)

Pau

t (m

m)

PAN

N (m

m)

Pw

in (m

m)

Mea

n te

mpe

ratu

re (deg

C)MTWA

MTCO

TANN

Age (cal yr BP)

Age (cal yr BP)

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

917-V

917-X

II

917-X

I

917-X

917-I

X

917-V

IIIb

917-V

I

917-I

V

917-I

II

917-I

917-I

I

917-V

IIIa

917-V

II

Pol

len

zona

tion

Present-day value




-40

-34

0

20

40

60

80

2000 4000 6000 8000 10000 12000

0

60

40

20

2000 4000 6000 8000 10000 12000

Age (cal yr BP)


Age (cal yr BP)

Semi desert ()

Temperate forest ()

-36

-38

-42

82

00 y

r eve

nt

PB

O

Youn

ger

Dry

as

PB

18

16

14

12

10









- 250

- 150

- 050

20

40

60

80

100

2000 4000 6000 8000 10000 12000 14000

050

150

250

Trees

18δ O permil

13Cδ C permil

Age (cal yr BP)

18

810121416
















1000 3000 1100050002000 4000 7000 80000 100009000 120006000

-4

-6

-8

-10

-12

2

0

-2

-4

-6MTW

A an

omal

ies

MTC

O a

nom

alie

s

MD 90-917

SW Europe

SW Europe

SE Europe

SE Europe

MD 90-917

1000 3000 1100050002000 4000 7000 80000 100009000 120006000

1

0

-1

-2

-3

2

1

0

-1

-2

0

-1

-2

-3

-4

3

2

1

0

Age (cal yr BP)

Age (cal yr BP)

MTW

A anomalies

MTC

O anom

aliesM

TWA anom

aliesM

TCO

anomalies






Combourieu-

YD YD

SL 152



200 40 60

Quercus ()

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age

(cal

yr B

P)

ACP2

ACP3

ACP4

ACP5

ACP6

ACP7

ACP8ACP9

YD

NAC1

NAC2

NAC3

NAC4

NAC5

NAC6

NAC7

NAC8

NAYD

MD 90-917


(2001)


cold events

YD

West-CentralEurope

YD

AdC1

AdC2

AdC4

AdC6

AdC7

AdC8

AdC3

AdC5

MD 01-2797

Desprat et al(2013)

MD 99-2043


Magny et al(2012)

ODP 976

Nebout et al(2009)

60


PBO

BO

8200

9200

4200

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age (cal yr BP)









2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

100

200Psum (mm)

S1aS1b

Age (cal yr BP)


+

-

2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

Age (cal yr BP)

-6


IK

Po

Riv

er

disc

harg

e

+

-

loca

l riv

ers

dis

char

ge

-

+

salin

e

+

- Sum

mer

Pre

cipi

tatio

n

-52

-49

-46

-43

-40

-37

15

20

1

2

3


- Pre

cipi

tatio

n

+

- Pre

cipi

tatio

n

3

5

7

9

Pwin (mm)100

200 +

- Win

ter

Pre

cipi

tatio

n

SK


PBOBO82 92 42

M

D

90

917





8 Conclusions











References

















































































































































0

10

20

30

40

50

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

0

10

20

30

400

500

600

700

800

900

1000

300

100

200100

200

100

200

300

100

200

60

80

70

0

10

20

30

40

50

60

80

70

0

10

20

0

10

20

30

0

10

20

30

40

0

10

Alti

tudi

nal

taxa

()

Oth

er te

mpe

rate

Cor

ylus

()

Tem

pera

te fo

rest

d

ecid

uous

Q

uerc

us (

)Med

iterr

anea

nta

xa (

)S

emi-d

eser

tta

xa (

)P

sum

(mm

)

Her

bace

ous

taxa

()

Psp

r (m

m)

Pau

t (m

m)

PAN

N (m

m)

Pw

in (m

m)

Mea

n te

mpe

ratu

re (deg

C)MTWA

MTCO

TANN

Age (cal yr BP)

Age (cal yr BP)

2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000

917-V

917-X

II

917-X

I

917-X

917-I

X

917-V

IIIb

917-V

I

917-I

V

917-I

II

917-I

917-I

I

917-V

IIIa

917-V

II

Pol

len

zona

tion

Present-day value




-40

-34

0

20

40

60

80

2000 4000 6000 8000 10000 12000

0

60

40

20

2000 4000 6000 8000 10000 12000

Age (cal yr BP)


Age (cal yr BP)

Semi desert ()

Temperate forest ()

-36

-38

-42

82

00 y

r eve

nt

PB

O

Youn

ger

Dry

as

PB

18

16

14

12

10









- 250

- 150

- 050

20

40

60

80

100

2000 4000 6000 8000 10000 12000 14000

050

150

250

Trees

18δ O permil

13Cδ C permil

Age (cal yr BP)

18

810121416
















1000 3000 1100050002000 4000 7000 80000 100009000 120006000

-4

-6

-8

-10

-12

2

0

-2

-4

-6MTW

A an

omal

ies

MTC

O a

nom

alie

s

MD 90-917

SW Europe

SW Europe

SE Europe

SE Europe

MD 90-917

1000 3000 1100050002000 4000 7000 80000 100009000 120006000

1

0

-1

-2

-3

2

1

0

-1

-2

0

-1

-2

-3

-4

3

2

1

0

Age (cal yr BP)

Age (cal yr BP)

MTW

A anomalies

MTC

O anom

aliesM

TWA anom

aliesM

TCO

anomalies






Combourieu-

YD YD

SL 152



200 40 60

Quercus ()

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age

(cal

yr B

P)

ACP2

ACP3

ACP4

ACP5

ACP6

ACP7

ACP8ACP9

YD

NAC1

NAC2

NAC3

NAC4

NAC5

NAC6

NAC7

NAC8

NAYD

MD 90-917


(2001)


cold events

YD

West-CentralEurope

YD

AdC1

AdC2

AdC4

AdC6

AdC7

AdC8

AdC3

AdC5

MD 01-2797

Desprat et al(2013)

MD 99-2043


Magny et al(2012)

ODP 976

Nebout et al(2009)

60


PBO

BO

8200

9200

4200

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age (cal yr BP)









2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

100

200Psum (mm)

S1aS1b

Age (cal yr BP)


+

-

2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

Age (cal yr BP)

-6


IK

Po

Riv

er

disc

harg

e

+

-

loca

l riv

ers

dis

char

ge

-

+

salin

e

+

- Sum

mer

Pre

cipi

tatio

n

-52

-49

-46

-43

-40

-37

15

20

1

2

3


- Pre

cipi

tatio

n

+

- Pre

cipi

tatio

n

3

5

7

9

Pwin (mm)100

200 +

- Win

ter

Pre

cipi

tatio

n

SK


PBOBO82 92 42

M

D

90

917





8 Conclusions











References




































































































































-40

-34

0

20

40

60

80

2000 4000 6000 8000 10000 12000

0

60

40

20

2000 4000 6000 8000 10000 12000

Age (cal yr BP)


Age (cal yr BP)

Semi desert ()

Temperate forest ()

-36

-38

-42

82

00 y

r eve

nt

PB

O

Youn

ger

Dry

as

PB

18

16

14

12

10









- 250

- 150

- 050

20

40

60

80

100

2000 4000 6000 8000 10000 12000 14000

050

150

250

Trees

18δ O permil

13Cδ C permil

Age (cal yr BP)

18

810121416
















1000 3000 1100050002000 4000 7000 80000 100009000 120006000

-4

-6

-8

-10

-12

2

0

-2

-4

-6MTW

A an

omal

ies

MTC

O a

nom

alie

s

MD 90-917

SW Europe

SW Europe

SE Europe

SE Europe

MD 90-917

1000 3000 1100050002000 4000 7000 80000 100009000 120006000

1

0

-1

-2

-3

2

1

0

-1

-2

0

-1

-2

-3

-4

3

2

1

0

Age (cal yr BP)

Age (cal yr BP)

MTW

A anomalies

MTC

O anom

aliesM

TWA anom

aliesM

TCO

anomalies






Combourieu-

YD YD

SL 152



200 40 60

Quercus ()

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age

(cal

yr B

P)

ACP2

ACP3

ACP4

ACP5

ACP6

ACP7

ACP8ACP9

YD

NAC1

NAC2

NAC3

NAC4

NAC5

NAC6

NAC7

NAC8

NAYD

MD 90-917


(2001)


cold events

YD

West-CentralEurope

YD

AdC1

AdC2

AdC4

AdC6

AdC7

AdC8

AdC3

AdC5

MD 01-2797

Desprat et al(2013)

MD 99-2043


Magny et al(2012)

ODP 976

Nebout et al(2009)

60


PBO

BO

8200

9200

4200

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age (cal yr BP)









2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

100

200Psum (mm)

S1aS1b

Age (cal yr BP)


+

-

2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

Age (cal yr BP)

-6


IK

Po

Riv

er

disc

harg

e

+

-

loca

l riv

ers

dis

char

ge

-

+

salin

e

+

- Sum

mer

Pre

cipi

tatio

n

-52

-49

-46

-43

-40

-37

15

20

1

2

3


- Pre

cipi

tatio

n

+

- Pre

cipi

tatio

n

3

5

7

9

Pwin (mm)100

200 +

- Win

ter

Pre

cipi

tatio

n

SK


PBOBO82 92 42

M

D

90

917





8 Conclusions











References




































































































































- 250

- 150

- 050

20

40

60

80

100

2000 4000 6000 8000 10000 12000 14000

050

150

250

Trees

18δ O permil

13Cδ C permil

Age (cal yr BP)

18

810121416
















1000 3000 1100050002000 4000 7000 80000 100009000 120006000

-4

-6

-8

-10

-12

2

0

-2

-4

-6MTW

A an

omal

ies

MTC

O a

nom

alie

s

MD 90-917

SW Europe

SW Europe

SE Europe

SE Europe

MD 90-917

1000 3000 1100050002000 4000 7000 80000 100009000 120006000

1

0

-1

-2

-3

2

1

0

-1

-2

0

-1

-2

-3

-4

3

2

1

0

Age (cal yr BP)

Age (cal yr BP)

MTW

A anomalies

MTC

O anom

aliesM

TWA anom

aliesM

TCO

anomalies






Combourieu-

YD YD

SL 152



200 40 60

Quercus ()

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age

(cal

yr B

P)

ACP2

ACP3

ACP4

ACP5

ACP6

ACP7

ACP8ACP9

YD

NAC1

NAC2

NAC3

NAC4

NAC5

NAC6

NAC7

NAC8

NAYD

MD 90-917


(2001)


cold events

YD

West-CentralEurope

YD

AdC1

AdC2

AdC4

AdC6

AdC7

AdC8

AdC3

AdC5

MD 01-2797

Desprat et al(2013)

MD 99-2043


Magny et al(2012)

ODP 976

Nebout et al(2009)

60


PBO

BO

8200

9200

4200

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age (cal yr BP)









2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

100

200Psum (mm)

S1aS1b

Age (cal yr BP)


+

-

2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

Age (cal yr BP)

-6


IK

Po

Riv

er

disc

harg

e

+

-

loca

l riv

ers

dis

char

ge

-

+

salin

e

+

- Sum

mer

Pre

cipi

tatio

n

-52

-49

-46

-43

-40

-37

15

20

1

2

3


- Pre

cipi

tatio

n

+

- Pre

cipi

tatio

n

3

5

7

9

Pwin (mm)100

200 +

- Win

ter

Pre

cipi

tatio

n

SK


PBOBO82 92 42

M

D

90

917





8 Conclusions











References








































































































































1000 3000 1100050002000 4000 7000 80000 100009000 120006000

-4

-6

-8

-10

-12

2

0

-2

-4

-6MTW

A an

omal

ies

MTC

O a

nom

alie

s

MD 90-917

SW Europe

SW Europe

SE Europe

SE Europe

MD 90-917

1000 3000 1100050002000 4000 7000 80000 100009000 120006000

1

0

-1

-2

-3

2

1

0

-1

-2

0

-1

-2

-3

-4

3

2

1

0

Age (cal yr BP)

Age (cal yr BP)

MTW

A anomalies

MTC

O anom

aliesM

TWA anom

aliesM

TCO

anomalies






Combourieu-

YD YD

SL 152



200 40 60

Quercus ()

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age

(cal

yr B

P)

ACP2

ACP3

ACP4

ACP5

ACP6

ACP7

ACP8ACP9

YD

NAC1

NAC2

NAC3

NAC4

NAC5

NAC6

NAC7

NAC8

NAYD

MD 90-917


(2001)


cold events

YD

West-CentralEurope

YD

AdC1

AdC2

AdC4

AdC6

AdC7

AdC8

AdC3

AdC5

MD 01-2797

Desprat et al(2013)

MD 99-2043


Magny et al(2012)

ODP 976

Nebout et al(2009)

60


PBO

BO

8200

9200

4200

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age (cal yr BP)









2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

100

200Psum (mm)

S1aS1b

Age (cal yr BP)


+

-

2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

Age (cal yr BP)

-6


IK

Po

Riv

er

disc

harg

e

+

-

loca

l riv

ers

dis

char

ge

-

+

salin

e

+

- Sum

mer

Pre

cipi

tatio

n

-52

-49

-46

-43

-40

-37

15

20

1

2

3


- Pre

cipi

tatio

n

+

- Pre

cipi

tatio

n

3

5

7

9

Pwin (mm)100

200 +

- Win

ter

Pre

cipi

tatio

n

SK


PBOBO82 92 42

M

D

90

917





8 Conclusions











References




































































































































Combourieu-

YD YD

SL 152



200 40 60

Quercus ()

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age

(cal

yr B

P)

ACP2

ACP3

ACP4

ACP5

ACP6

ACP7

ACP8ACP9

YD

NAC1

NAC2

NAC3

NAC4

NAC5

NAC6

NAC7

NAC8

NAYD

MD 90-917


(2001)


cold events

YD

West-CentralEurope

YD

AdC1

AdC2

AdC4

AdC6

AdC7

AdC8

AdC3

AdC5

MD 01-2797

Desprat et al(2013)

MD 99-2043


Magny et al(2012)

ODP 976

Nebout et al(2009)

60


PBO

BO

8200

9200

4200

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

Age (cal yr BP)









2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

100

200Psum (mm)

S1aS1b

Age (cal yr BP)


+

-

2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

Age (cal yr BP)

-6


IK

Po

Riv

er

disc

harg

e

+

-

loca

l riv

ers

dis

char

ge

-

+

salin

e

+

- Sum

mer

Pre

cipi

tatio

n

-52

-49

-46

-43

-40

-37

15

20

1

2

3


- Pre

cipi

tatio

n

+

- Pre

cipi

tatio

n

3

5

7

9

Pwin (mm)100

200 +

- Win

ter

Pre

cipi

tatio

n

SK


PBOBO82 92 42

M

D

90

917





8 Conclusions











References




































































































































2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

100

200Psum (mm)

S1aS1b

Age (cal yr BP)


+

-

2000 3000 4000 5000 6000 7000 8000 9000 100001100012000

Age (cal yr BP)

-6


IK

Po

Riv

er

disc

harg

e

+

-

loca

l riv

ers

dis

char

ge

-

+

salin

e

+

- Sum

mer

Pre

cipi

tatio

n

-52

-49

-46

-43

-40

-37

15

20

1

2

3


- Pre

cipi

tatio

n

+

- Pre

cipi

tatio

n

3

5

7

9

Pwin (mm)100

200 +

- Win

ter

Pre

cipi

tatio

n

SK


PBOBO82 92 42

M

D

90

917





8 Conclusions











References










































































































































References

































































































































