ORIGINAL

Late Glacial–Holocene ecostratigraphy of the south-easternAegean Sea, based on plankton and pollen assemblages

M. V. Triantaphyllou & A. Antonarakou & K. Kouli &M. Dimiza & G. Kontakiotis & M. D. Papanikolaou &

P. Ziveri & P. G. Mortyn & V. Lianou & V. Lykousis &

M. D. Dermitzakis

Received: 15 October 2008 /Accepted: 30 March 2009 /Published online: 16 April 2009# Springer-Verlag 2009

Abstract Quantitative analyses of coccolithophores, plank-tonic foraminifers, dinoflagellate cysts and pollen assemb-lages were carried out on shallow (NS-14) and deeper (NS-40)sediment cores from the south-eastern Aegean Sea. Ninecoccolithophore (ACE 1–9) and nine planktonic foraminifer(APFE 1–9) ecozones, correlated with dinoflagellate cystevidence, have been defined for the last ~14.5 cal. ka.Additionally, eight pollen assemblage zones (PAZ 1–8) have

been recognised and correlated with the plankton ecozones.Although generally consistent with existing schemes for thecentral and eastern Mediterranean, the established high-resolution ecostratigraphy has led to an expanded palae-oecological reconstruction of the Late Glacial–Holocenearchive in the south-eastern Aegean Sea, defining two warmand humid phases at 9.3–8.6 and 7.6–6.4 cal. ka B.P.,associated with the deposition of the early Holocene sapropelS1, and a third one between 5.2 and 4.2 cal. ka B.P. The highsedimentation rates which characterise the study area enabledthe detection of even minor and brief climatic events in theAegean Sea during S1 deposition times.

Introduction

In contrast to modern environmental conditions of lowbiological production and oxygen-replete bottom waters inthe eastern Mediterranean Sea, the presence of organic-richsapropel layers demonstrates that dramatically differentconditions have occurred periodically in the entire easternMediterranean since the Pliocene. These unusual depositscoincide with insolation maxima which caused a wetter andwarmer climate in the region (Rossignol-Strick 1985; Lourenset al. 1992), due to intensification of the monsoon system(e.g. Rohling et al. 2002a) and/or water circulation andbiogeochemical cycling (Anastasakis and Stanley 1986; vanSantvoort et al. 1997). In the sea, convection and mixing wasimpeded due to warming and freshening of the surface waterlayer during sapropel times (e.g. Emeis et al. 2000, 2003).Export productivity from the pelagic layer increased(Mercone et al. 2001), while preservation of organic matterimproved under oxygen-deficient bottom-water conditions(Moodley et al. 2005). The early Holocene sapropel S1 wasdeposited in the North Aegean Sea between 9.6 and 6.7 cal.

Geo-Mar Lett (2009) 29:249–267DOI 10.1007/s00367-009-0139-5

Electronic supplementary material The online version of this article(doi:10.1007/s00367-009-0139-5) contains supplementary material,which is available to authorized users.

M. V. Triantaphyllou (*) :A. Antonarakou :K. Kouli :M. Dimiza :G. Kontakiotis :V. Lianou :M. D. DermitzakisFaculty of Geology and Geoenvironment,Department of Historical Geology–Paleontology,National and Kapodistrian University of Athens,Panepistimiopolis, 15784 Athens, Greecee-mail: mtriant@geol.uoa.gr

M. D. PapanikolaouDepartment of Geography,University of Cambridge, Cambridge Quaternary,CB2 3EN, Cambridge, UK

P. ZiveriDepartment of Paleoclimatology and Geomorphology, FALW,Vrije Universiteit Amsterdam,1081 HVAmsterdam, The Netherlands

P. Ziveri : P. G. MortynICTA, Autonomous University of Barcelona (UAB),Edifici Cno-Campus UAB,08193 Bellaterra, Spain

V. LykousisHellenic Centre for Marine Research,Institute of Oceanography,19013 Anavyssos, Greece

ka B.P. (Perissoratis and Piper 1992), in the south-westernAegean Myrtoon Basin at 9.0–6.9 uncal. ka B.P. (Geragaet al. 2000), in the South Aegean at 9.8–6.3 cal. ka B.P.

(Rohling et al. 2002b) and in the central Aegean at 9.0–6.0 cal. ka B.P. (Geraga et al. 2005). Interruptions withinsapropel S1, centred roughly at 7.5 uncal. ka B.P., have beendocumented in the Adriatic (Rohling et al. 1997) and Aegeanseas (Perissoratis and Piper 1992; De Rijk et al. 1999;Geraga et al. 2000), and have unambiguously been inter-preted as real events of bottom-water re-oxygenation relatedto cooling episodes (De Rijk et al. 1999).

Palaeoceanographic studies (Casford et al. 2002; Rohlinget al. 2002b, 2006; Marino 2007; Gogou et al. 2007a) havedemonstrated that the Aegean Sea sediment record providesdirect evidence of high-latitude climate influences. Thisgeneral response of the eastern Mediterranean Sea to globalclimate change is expressed differently in sediment recordsfrom the various sub-basins. Therefore, numerous workshave established ecostratigraphic schemes based on plank-tonic foraminifers, coccolithophores, dinoflagellates, benthicforaminifers, pollen, as well as stable isotope data (e.g.Jorissen et al. 1993; Capotondi et al. 1999; Sbaffi et al. 2001,2004; Giunta et al. 2003; Principato et al. 2003; Geraga et al.2005, 2008), and have shown the existence of palaeoenvir-onmental changes corresponding to major and minor climaticfluctuations during the Late Glacial–Holocene in the central–eastern Mediterranean area.

Within the context presented above, the main goal of thisstudy is the establishment of a detailed ecostratigraphic scheme

for the south-eastern Aegean region, based on coccolitho-phore, planktonic foraminifer and dinoflagellate cyst patternsas well as pollen assemblages characterising the last ~14.5 cal.ka B.P. In addition, a detailed palaeoecological reconstructionof the Late Glacial–Holocene interval is presented, emphasis-ing sapropel S1 formation in both shallow and deeperarchives along the south-eastern margin of the Aegean Sea.Intercorrelation of our marine and terrestrial data from theAegean Sea with the existing schemes has enabled an evenhigher-resolution evaluation of recorded climatic changeswithin the eastern Mediterranean.

Oceanographic setting

The semi-enclosed Aegean Sea is ideal for reconstructionsof past climatic changes, due to its unique physical andgeographic configuration displaying complicated seabedmorphology and numerous island complexes. It is situatedbetween Turkey and Greece (Fig. 1a), and is linked to theBlack Sea through the Dardanelles Strait and to the easternMediterranean through the Cretan Straits. More impor-tantly, the Aegean Sea hosts the main source area of newdeepwater formation for the entire eastern Mediterranean(Theocharis et al. 1999; Zervakis et al. 2000).

The cooler (9–22°C) and fresher (22–33‰) Black Seasurface waters flow to the southwest Aegean Sea along theeast coast of Greece, enhancing the productivity of the NorthAegean Sea. Freshwater influx into the Aegean Basin comes

Fig. 1 a Location of the study area in the SE Aegean Sea andprincipal circulation patterns of the surface waters, in winter (openarrows) and summer (closed arrows). Arrowed ovals denote cyclones

(based on data from Lacombe and Tchernia 1972; Theocharis andGeorgopoulos 1993; Poulos et al. 1997). b Locations of cores NS-14and NS-40

250 Geo-Mar Lett (2009) 29:249–267

also from the discharge of several major rivers of the northAegean coasts and the east Turkish coast (Poulos et al. 1997).Warm (16–25°C) and highly saline (39.2–39.5‰) easternMediterranean waters flow northwards along the easternAegean as far as the Dardanelles Strait (Zervakis et al. 2000).

The Aegean Sea is one of the most oligotrophic areas in theeastern Mediterranean Sea, itself amongst the most nutrient-limited regions of the global ocean system (Ignatiades et al.2002; Lykousis et al. 2002). The South Aegean sub-basin inparticular is considered as a "typical oceanic margin"environment (Lykousis et al. 2002), characterised by verylow export rates of organic carbon from the surface layer, andorganic-poor sediments (Gogou et al. 2000); coccolitho-phores are one of the major primary producers in thisecosystem (Ignatiades et al. 2002; Triantaphyllou et al. 2004).Emiliania huxleyi dominates the winter coccolithophoreassemblages, followed in summer by holococcolithophoresand other species such as Rhabdosphaera clavigera andSyracosphaera spp. (Dimiza et al. 2008). Planktonic fora-minifer assemblages are dominated during the summer periodby both varieties of Globigerinoides ruber (G. ruber albaand G. ruber rosea), whereas G. ruber alba and Globigerinabulloides are present mostly in winter (Pujol and VergnaudGrazzini 1995).

The area lying between the islands of Kos and Nisyros(Fig. 1b) represents the easternmost edge of the volcanicarc characterising the recent Aegean Sea. The prominenttectonic features of the area, namely the active fault zone ofKos Island, reflect the complexity of the bottom morphol-ogy and the existence of five distinct marine basins,including the western Kos Basin (location of one of our

two cores, NS-14; Fig. 1b). The depth of this basin does notexceed 510 m (Papanikolaou and Nomikou 2001). Mainsources of material influx for the basin are the erodingvolcanic rocks from the surrounding islands, and particulatematter discharge from the rivers of the Turkish coast.

Study site and age model

Gravity cores were recovered during the R/V Aegeo cruiseof 1998, from a shallower (NS-14) and a deeper (NS-40)site at the easternmost edge of the recent Aegean volcanicarc (Fig. 1, Table 1).

The dominant lithology of core sediments is greyhemipelagic mud, intercalated with tephra, sapropel andturbidite layers (Triantaphyllou et al. 2006, 2007). In coreNS-14, the most recent Z2 Santorini tephra layer occurs at17 cm depth. This is followed by a dark layer at 40–25 cmdepth, with higher organic content (Gogou et al. 2007b).Sapropel S1 can be identified slightly deeper, from 55 to120 cm, and is distinctly subdivided into two units (S1a,S1b) by a lighter grey-coloured interval at 69–80 cm. Aturbiditic layer is present between 231 and 240 cm, whereasgrey clay with pebbles prevail from 300 cm to the corebottom, representing a gravity flow event. Samples formicropalaeontological analysis were collected only fromthe upper 300 cm of core NS-14.

In core NS-40, the Z2 tephra layer appears at a depth of38 cm, and S1 between 66 and 42 cm. The latter layer canbe subdivided into units S1a and S1b, with an interruptioninterval between 49 and 52 cm. For the needs of the present

Table 1 Position and data used for the construction of the age model: cores NS-14 and NS-40 (cf. Triantaphyllou et al. 2006, 2007)

Depth (cm) Age, non-calibrated (ka B.P.) Age, calibrated (ka B.P.) Age source

NS-14: 36°38.92′ N, 27°00.47′ E, 505 m water depth

0 2.464 Core top

17 3.581 Santorini ash layer Z2 (Ramsey et al. 2004)

48 5.490±40 5.587–5.853a AMS14C

80 8.280±50 8.430–8.843a AMS14C

86 8.390±40 8.596–8.959a AMS14C

114 8.640±40 8.968–9.299a AMS14C

252 11.770±60 12.956–13.240a AMS14C

NS-40: 36°19.04′ N, 27°01.13′ E, 1,078 m water depth

38 3.581 Santorini ash layer Z2 (Ramsey et al. 2004)

42 6.100 6.353 Top S1bb

49 7.320 7.622 Bottom S1bb

52 8.280 8.636 Top S1ab

66 8.776 9.316 Bottom S1ab

a Calibrated ages were obtained with the CALIB 5.01 program, with a regional reservoir age correction of 149±30 years for the Aegean (Facorelliset al. 1998)b Ages based on NS-14 dating

Geo-Mar Lett (2009) 29:249–267 251

study, micropalaeontological samples were collected onlyfrom the upper 100 cm of core NS-40.

The age model for core NS-14 is based on linearinterpolation between five AMS datings (see Table 1;Triantaphyllou et al. 2006). According to this model, thebase of S1a has an estimated age of 9.3 cal. ka B.P. and thetop of S1a one of 8.6 cal. ka B.P. The top of the S1interruption is dated at 7.6 cal. ka B.P. and the top of S1b at6.4 cal. ka B.P. The corresponding average sedimentationrate is 57 cm 1,000 years−1 in sapropel layer S1a. In thesapropel interruption interval, the values decrease stronglyto 11 cm 1,000 years−1, and this low rate persists up to theupper S1b sapropel.

The age model for core NS-40 is based on correlation withthe top and bottom of S1 in core NS-14 (Table 1), as bothcores were recovered at proximate sites and have parallelstratigraphies. Average sedimentation rates for core NS-40are 21 cm 1,000 years−1 for S1a, 3 cm 1,000 years−1 for theinterruption interval, and 6 cm 1,000 years−1 for S1b.

Materials and methods

Subsamples for micropalaeontological analysis were takendowncore at about 2-cm intervals, except for the sapropellayers where the sampling resolution was 1 cm. Coccolitho-phore analyses were carried out on 156 samples from coreNS-14 and 31 samples from core NS-40. Sample preparationwas according to standard smear slide techniques. Analyseswere performed using a Leica DMSP optical polarising lightmicroscope at ×1,250 magnification, by counting at least 300specimens per sample. Specimens resembling Reticulofenes-tra spp. were assigned to the Emiliania huxleyi moderatelycalcified (EHMC) morphotypes of Crudeli et al. (2004,2006). Additional counts of 15 fields of view (Negri andGiunta 2001) were performed for Helicosphaera spp.,Rhabdosphaera spp., Syracosphaera spp., Coronosphaeraspp. and Calcidiscus spp. (field of view area 0.02 mm2).The rare species Braarudosphaera bigelowii was countedin a fixed area of 150 fields of view. All results wereconverted into relative abundances (percentages), in orderto avoid any dilution effects of, e.g. terrigenous matterinput (Flores et al. 1997).

Foraminifer analyses were carried out on 67 samples fromcore NS-14 and 55 samples from core NS-40. The sedimentwas washed through a 150-µm mesh sieve and dried at 60°C.For each sample, quantitative analysis was carried out onaliquots collected by means of a microsplitter, on at least 300specimens of planktonic foraminifers which were identifiedand classified following the taxonomic concept of Hemlebenet al. (1989). Neogloboquadrina pachyderma right- and left-coiling (dextral and sinistral), and the two varieties Globi-gerinoides ruber alba and rosea were counted and plotted

separately. Globigerinoides sacculifer includes also Globi-gerinoides trilobus, according to Hemlemben et al. (Hemle-ben et al., 1989). The species Globoturborotalita rubescensand Globoturborotalita tenella were counted and plottedtogether (Capotondi et al. 1999). Identified species areexpressed as percentages (%) of the total number ofplanktonic Foraminifera.

The palaeoclimatic curve inferred from the planktonicforaminifer assemblages was obtained by the formula (w–c)/(w + c) × 100, where w represents the warm-water indicatorsand c the cold-water indicators. Globigerinoides sacculifer,Globigerinoides ruber var. alba and rosea, Globoturborota-lita rubescens, Globorotalia truncatulinoides, Orbulinauniversa and Globigerinella siphonifera were consideredwarm-water indicators, and Globorotalia scitula, Globorota-lia inflata, Turborotalita quinqueloba and Globigerinitaglutinata cold-water indicators (Hemleben et al. 1989;Rohling et al. 1993; Pujol and Vergnaud Grazzini 1995).

Sample processing for dinoflagellate cyst analyses includeddemineralisation in cold 10% HCl to remove carbonates, andin 38%HF (in some cases, 48%HF) to remove silicates. Everytreatment was followed by washing with de-ionised water. Theresulting neutralised residue was sieved at 10 µm using a nylonmesh. One Lycopodium spore tablet was added to eachsample during treatment, to ensure estimation of absoluteabundances and reliability of quantitative data. Dinocystanalysis was performed on 18 samples, corresponding to thedepth interval 10–130 cm in core NS-14, with a Leica DMLBmicroscope. Acritarchs and pollen were not counted.

Because dinoflagellate cysts were rare (30–50 specimensper slide), and occasionally poorly preserved, palaeoenvi-ronmental interpretation of dinocyst abundances is somewhatrestricted in the present case. Nevertheless, heterotrophicspecies susceptible to oxidation (e.g. Stelladinium spp.,Selenopemphix spp., S. nephroides, S. quanta, Lejeunecystaspp.; Zonneveld et al. 2001, 2007) are indeed present, andtheir fluctuations in relative abundance largely follow thoseof autotrophs (e.g. Spiniferites mirabilis, S. hyperacanthus,Lingulodinium machaerophorum, Nematosphaeropsis laby-rinthus, Operculodinium centrocarpum sensu Wall & Dale,O. israelianum, Polysphaeridium zoharyi, P. reticulata,Impagidinium aculeatum, I. patulum, I. paradoxum, Penta-pharsodinium dalei). This suggests that oxidation effects arenon-significant and likely provides evidence that heterotrophincrease reflects higher palaeoproductivity, rather than onlygood preservation. Therefore, a tentative estimation ofpalaeoproductivity has been made based on the heterotrophic(Protoperidinoid + T. vancampoae)/(autotrophic) (Gonyau-lacoid) ratio (H/A ratio).

Pollen analyses were performed on 134 weighed samplesfrom core NS-14; diagrams were not constructed from210 cm downwards, due to extremely low pollen contents.Samples were treated with 10% HCl and 38% HF, and

252 Geo-Mar Lett (2009) 29:249–267

sieved through a 10-µm sieve; Lycopodium spores wereadded in order to check quantitative data reliability. Thesamples were stored and mounted in silicone oil, and aminimum of 150 pollen grains was counted in each case.Identifications were aided by the keys and descriptions ofMoore et al. (1991) and Reille (1992), and examination ofreference grains. Pinus, pteridophyte spores and aquatictaxa were excluded from pollen budgets. Relative abun-dances (percentages) and contents (grains per gram of drybulk sediment; cf. Flemming and Delafontaine 2000) werecalculated for selected pollen types. Contents of palyno-morph taxa in rivers and lakes (like Sparganium, Pedias-trum and Zygnemantaceae) have been used as proxies forriver runoff into the Aegean Basin (Targarona 1997).

Taxonomic lists of the various taxa and species identified/counted can be found online in the electronic supplementarymaterial for this study.

Results

Coccolithophores

In core NS-14, Emiliania huxleyi abundances remain consis-tently above 60% from 300 up to 135 cm depth (Fig. 2).Above this, its relative frequency decreases, to a minimumvalue of 25% during the S1 interruption interval. E. huxleyiincreases again from 30 cm towards the core top. In core NS-40, E. huxleyi abundances present a gradual decrease from75 cm to the base of S1a (Fig. 3). During S1 deposition,mean values average at 35%, with a rather abrupt increaseprior to the top S1b at 45 cm. Emiliania huxleyi moderatelycalcified morphotypes (EHMC; see Fig. 2) are found mainlyin the interval from 300 to 120 cm, below S1a, and reappearat 60 cm at the top of S1b, with a maximum abundance at15 cm from the NS-14 core top. In core NS-40, EHMCmorphotypes are almost absent in S1a, reappearing with verylow values in the S1b interval (Fig. 3). In the latter core,these morphotypes increase abruptly at 40 cm, above S1.

In core NS-14, Florisphaera profunda starts to increase inabundance below S1 at 135 cm. High abundances (approx.40%) are reached in the upper part of S1a and, surprisingly,this continues throughout the S1 interruption between 80 and69 cm. A sharp decrease of F. profunda is recorded at 60 cm,still within the S1b interval (Fig. 2). A prominent peak inabundance (up to 52%) is observed at 32 cm (NS-14). In coreNS-40, F. profunda starts to increase below S1 at 76 cm,reaching 50% frequency at the beginning of S1a (Fig. 3). Inboth cores, F. profunda generally presents an inverse patternwith respect to E. huxleyi, and dominates the coccolithophoreassemblages during the S1 deposition interval.

In core NS-14, Helicosphaera spp. (mainly H. carteri)starts to increase at 135 cm, peaking in the lower part of

S1a, between 120 and 100 cm (Fig. 2). Another intervalwith high frequencies of Helicosphaera spp. and Syracos-phaera spp. (mainly S. pulchra) occurs between 40–25 cmin core NS-14. In core NS-40, Helicosphaera spp.,Syracosphaera spp. and Rhabdosphaera spp. presentsimilar distribution patterns from the base of S1a up to56 cm, with a frequency peak at 63 cm depth (Fig. 3).

Gephyrocapsa oceanica occurs mainly between 60 and29 cm, with maximum abundances in the upper part of S1bin core NS-14; outside this interval, its presence is onlysporadic. Calcidiscus spp. and Umbilicosphaera spp.increase in abundance above 100 cm but the highestfrequencies are recorded above S1. Umbellosphaera tenuispresents higher values between 300 and 260 cm and aboveS1 up to 29 cm in core NS-14, whereas this species shows agradual increase close to the base of S1a in core NS-40.Braarudosphaera bigelowii is present at 170 and at 78 cm,but increases abruptly in the upper part of S1b at 57–55 cmdepth in NS-14; it still occurs (albeit with very fewrepresentatives) up to 29 cm (Fig. 2). In core NS-40, B.bigelowii is present only for small intervals in S1a, butincreases abruptly at the top of S1a, during the S1interruption and at the top of S1b.

Planktonic foraminifers

Globigerinoides ruber, represented by two varieties, G.ruber alba (white) and G. ruber rosea (pink), is the mostabundant species in both cores, displaying maximum values(up to 60%) within sapropel S1 (Figs. 4, 5). In particular, incore NS-14 after a decrease (from 60 to 25%) at the base ofthe core, G. ruber alba gradually increases to reachmaximum abundance within sapropel S1a (Fig. 4). Thespecies decreases during the interruption of S1, and reachesa maximum again between 40–25 cm depth and at the topof the core. G. ruber rosea first appears at the beginning ofS1 and reaches maximum abundance within S1a, at the topof S1b and between 40–25 cm depth.

In core NS-40, G. ruber alba reaches maximum abun-dance immediately below the deposition of, and within thelower part of sapropel S1a (Fig. 5). High abundances are alsorecorded at the top of S1a and S1b, following a slightdecrease at 57 cm (middle S1a) and 47 cm (lower S1b).G. ruber alba clearly decreases during the interruption. Bycontrast, G. ruber rosea is always high in abundance, andreaches maximum values in the lower part of S1a and S1b.

The relative abundance curve of G. sacculifer displays ahighly variable downcore pattern, occasionally reachingsignificant percentages. The appearance of this species atthe base of S1a is pronounced in both cores. It disappearstemporarily in S1a but again displays an increasing trend inthe upper part of S1b, and within the dark layer between40–25 cm depth in core NS-14.

Geo-Mar Lett (2009) 29:249–267 253

Fig. 2 Relative abundance (%) of coccolithophore species for coreNS-14. To the right, the established Aegean coccolithophore ecozones(ACE). EHMC denotes the Emiliania huxleyi moderately calcified

morphotype. NS-14 lithology: Z2 Santorini ash layer, D dark layerwith higher organic content, S1a, S1b sapropel layers, T turbidite layer

254 Geo-Mar Lett (2009) 29:249–267

Globigerina bulloides is continuously present, display-ing almost the same distributional pattern in both cores NS-14 and NS-40 (Figs. 4, 5). The species increases withinS1a, and abruptly decreases at the top of S1a, S1b and at25 cm depth in core NS-14, as well as in the upper part ofS1b in core NS-40.

Neogloboquadrinids represented by Neogloboquadrinapachyderma and N. dutertrei are absent in the sapropelintervals. N. dutertrei occurs only in the lower part of coreNS-14, whereas N. pachyderma is present below thedeposition of S1a, within the interruption and above thesapropel interval.

Globigerinita glutinata displays high abundances belowthe deposition of S1, and decreases in the sapropel intervalsin both cores (but more markedly in core NS-40; Fig. 5).An increase of the species is evident at the top of S1b inboth cores.

Turborotalita quinqueloba is discontinuously present inboth cores, occurring mostly below and above sapropeldeposition and within the interruption.

Among the minor species, which display discontinuousdistributions, G. scitula is present only in core NS-14,restricted to the lower part of the core (Fig. 4). In both coresthe warm-water indicators G. rubescens and G. siphonifera

Fig. 4 Relative abundance (%) of planktonic foraminifer species for core NS-14. To the right, the calculated palaeoclimatic curve and theestablished Aegean planktonic foraminifer ecozones (APFE)

Fig. 3 Relative abundance (%) of coccolithophore species for core NS-40. To the right, the established Aegean coccolithophore ecozones (ACE).NS-40 lithology: Z2 Santorini ash layer, S1a, S1b sapropel layers

Geo-Mar Lett (2009) 29:249–267 255

display a similar, overall gradually increasing trend withinS1a, whereas O. universa shows several peaks in abundanceduring the deposition of S1a, S1b and between 40–25 cm incore NS-14. Globorotalia truncatulinoides and Globorotaliainflata both disappear in the lower part of S1. The latterspecies reappears temporarily within the interruption andabove S1b in both study cores (Figs. 4, 5).

Dinoflagellate cysts

The dinocyst record for core NS-14 reveals overall lowabundances. Both auto- and heterotrophic species are wellrepresented. Variations in heterotrophic and autotrophicspecies follow a similar pattern, and slight indications ofdegradation (e.g. thin, pale brown or bleached walls) of theheterotrophic cysts were seen under the microscope. Giventhe fact that the heterotrophic component is rather wellrepresented, it is suggested that any degradation ofheterotrophic taxa susceptible to oxidation is likely insig-nificant (also cf. above). The S1a, S1b and the intervalbetween 40 and 25 cm depth are characterised by increasedrelative percentages of heterotrophs, but the highest valuesare clearly seen in the lower part of the S1a sapropel(Fig. 6). Apart from the heterotrophs, there are also a fewautotrophic taxa which constitute a prominent proportion ofthe cyst association in the sapropel intervals. Thus,Impagidinium aculeatum is well represented in S1b andparticularly S1a (5.8 and 13.6% respectively). Pyxidinopsisreticulata reaches 6.2% in the upper S1a sapropel.Lingulodinium machaerophorum (12%) and Polysphaeri-dium zoharyi subsp. ktana (15%) have high values belowS1. L. machaerophorum is particularly abundant in S1b(14.6%), less so in S1a (6.4%) and far less between 40 and25 cm depth. Nematosphaeropsis labyrinthus representsanother autotrophic taxon forming part of the sapropel

dinocyst assemblages in core NS-14, except for the intervalbetween 40 and 25 cm depth.

Pollen

Pollen contents in core NS-14 range from 10 to 53,000 grainsper gram dry sediment, the higher values being observed inthe sapropel layers S1a, S1b, and between 40 and 25 cm depth(Fig. 7). The pollen spectra reflect diversified vegetationranging from steppe to semi-desert (Artemisia, Chenopodia-ceae), and to a rich Mediterranean vegetation with evergreenelements and high herb diversity (Olea, Pistacia, evergreenQuercus, Helianthemum, Sanguisorba), as well as moun-tainous deciduous (Quercus cerris type, Acer, Ostrya) andconiferous forest representatives (Pinus, Abies, Cedrus).

Arboreal taxa are continuously present in all pollen spectra,though their abundances fluctuate. Below S1a, they representca. 25% of the pollen sum. Deciduous and evergreen taxa, suchas Quercus, Pistacia and Olea, increase at the base of S1a.Quercus (both evergreen and deciduous) shows its highestabundance between 120 and 29 cm; values are considerablylower outside the sapropel layers (Fig. 7). Cedrus isconsistently present up to 25 cm from the core top, exceptfor the interval of S1 interruption. Abies and Picea appear lessabundantly, mainly above 70 cm. Pistacia is nearly alwayspresent from the base of S1 (120 cm) upwards, with aprominent peak at 25 cm, being absent for short intervalsduring the S1 interruption, and above the deposition of S1.Steppe vegetation elements, such as Asteraceae, Poaceae andChenopodiaceae, occur in high numbers below S1. TheChenopodiaceae curve shows a pronounced maximum at135–120 cm below the base of S1 (Fig. 7). The semi-deserttaxa Artemisia and Ephedra are almost continuously presentat least in low percentages, and show relatively highabundances from 135 cm downwards (Fig. 7).

Fig. 5 Relative abundance (%) of planktonic foraminifer species for core NS-40. To the right, the calculated palaeoclimatic curve and theestablished Aegean planktonic foraminifer ecozones (APFE)

256 Geo-Mar Lett (2009) 29:249–267

The contents of aquatic palynomorphs (e.g. Sparganium,Pediastrum and Zygnemantaceae) are distinctly increasedwithin the sapropel layers and between 40 and 25 cm depth.

Ecostratigraphy

Based on the detailed analyses reported above for coccolitho-phore, planktonic foraminifer, dinoflagellate and pollenabundance in the shallow NS-14 and deeper NS-40 south-eastern Aegean cores, we present Aegean-specific ecostrati-graphic schemes in Fig. 8, together with their equivalentintervals in existing regional plankton ecostratigraphicschemes for the Tyrrhenian Sea (Sbaffi et al. 2001, 2004),eastern Mediterranean (Principato et al. 2003), Adriatic Sea(Jorissen et al. 1993; Capotondi et al. 1999; Giunta et al.2003) and Ionian Sea (Geraga et al. 2008), in particularduring the interval of sapropel S1 deposition. Ecozones havebeen defined in terms of fluctuations in abundance,frequency peaks and/or local occurrences or temporary

disappearances of specific coccolithophore, planktonic fora-minifer and pollen taxa selected on the basis of their present-day ecological significance and habitat preferences. In thecase of planktonic foraminifers, comparison between thedifferent ecozones is based on counts of specimens sized>150 µm, except for those of Capotondi et al. (1999) whoused the >63 µm fraction. This difference would probablylead to an over- or underestimation in the percentage of smallspecies such as T. quinqueloba and G. glutinata and of largerspecies such as G. sacculifer (Capotondi et al. 2004; Geragaet al. 2008). However, any such discrepancies appear to benegligible because our ecozonal scheme matches well withthat of Capotondi et al. (1999). In addition, Capotondi et al.(2006) have shown that the stratigraphic position of faunalevents does not essentially change when comparing differentsize fractions during intervals of similar environmentalconditions. Minor discrepancies between equivalent ageintervals are attributed mainly to differences in age modelcalculations in the various studies.

Fig. 6 Contents of total dinoflagellates and relative abundance (%) of the most indicative dinoflagellate taxa for core NS-14. H/A index denotesthe heterotrophic (Protoperidinoid + T. vancampoae)/(autotrophic) (Gonyaulacoid) ratio (see main text for further explanation)

Geo-Mar Lett (2009) 29:249–267 257

Coccolithophore ecozones

Nine Aegean coccolithophore ecozones (ACE-9 to ACE-1)are recognised since the Late Glacial (Fig. 8).

– ACE-9 (~14.5–13.1 cal. ka B.P.). Top: increase of R.clavigera. Interval: increase of E. huxleyi (up to 80%)and U. tenuis (1%). Equivalence: upper parts of ecozoneC6 of Principato et al. (2003) and of interval 5 of Giuntaet al. (2003), biozone 5 of Sbaffi et al. (2001).

– ACE-8 (13.1–9.9 cal. ka B.P.). Top: peak in Syracos-phaera spp., increase of F. profunda (30%). Interval:high abundance of E. huxleyi, decreased values ofHelicosphaera spp., peak of EHCM (approx. 6%).

Equivalence: ecozone C5 of Principato et al. (2003),interval 4 of Giunta et al. (2003), biozones 3, 4 of Sbaffiet al. (2001).

– ACE-7 (9.9–9.3 cal. ka B.P.). Top: increase of Helicos-phaera spp. (approx. 1%), decrease of EHCM. Interval:gradual increase of F. profunda and Helicosphaera spp.High abundances of dinocysts L. machaerophorum, N.labyrinthus and I. sphaericum. Equivalence: lower partof ecozone C4 of Principato et al. (2003), lower part ofbiozone 2 of Sbaffi et al. (2001).

– ACE-6b (9.3–9.0 cal. ka B.P.). Top: increase of Umbil-icosphaera spp., decrease of Helicosphaera spp. Interval:increased Helicosphaera spp. (up to 3%), and high valuesof R. clavigera and the dinoflagellate I. aculeatum.

Fig. 7 Relative abundance (%) of the most indicative pollen taxa forcore NS-14. Other deciduous species include Acer, Corylus, Carpinus,Alnus, Ulmus and Juglans. Aquatic palynomorphs include Sparganium,

Typha, Lemna, Potamogeton, Nymphaea, Pediastrum and Zygnemata-ceae. To the right, the established pollen assemblage zones (PAZ)

258 Geo-Mar Lett (2009) 29:249–267

EHCM morphotypes are absent. Equivalence: middle partof ecozone C4 of Principato et al. (2003), lower part ofbiozone 2 of Sbaffi et al. (2001).

– ACE-6a (9.0–8.8 cal. ka B.P.). Top: peak in F. profunda(approx. 60%), decrease of R. clavigera. Interval: highabundance of F. profunda, E. huxleyi and the dinofla-gellate P. reticulata. EHCM morphotypes are absent.Equivalence: upper part of ecozone C4 of Principatoet al. (2003), lower part of biozone 2 of Sbaffi et al.(2001).

– ACE-5 (8.8–7.6 cal. ka B.P.). Top: increase of F. profunda,E. huxleyi, Syracosphaera spp. Interval: presence of B.bigelowii. Equivalence: lower part of ecozone C3 ofPrincipato et al. (2003), interval 3 of Giunta et al. (2003),middle part of biozone 2 of Sbaffi et al. (2001).

– ACE-4b (7.6–6.9 cal. ka B.P.). Top: abrupt decrease inabundance of F. profunda (20%), decrease of Syracos-phaera spp. Interval: peak abundance of Rhabdos-phaera spp., Umbilicosphaera spp., Syracosphaeraspp., presence of B. bigelowii. Equivalence: middlepart of ecozone C3 of Principato et al. (2003), interval3 of Giunta et al. (2003), upper part of biozone 2 ofSbaffi et al. (2001).

– ACE-4a (6.9–6.4 cal. ka B.P.). Top: decrease ofHelicosphaera spp., distinct reduction of B. bigelowii,peak of G. oceanica. Interval: low abundance ofHelicosphaera spp. and F. profunda, peak of B.bigelowii. Equivalence: upper part of ecozone C3 ofPrincipato et al. (2003), interval 3 of Giunta et al.(2003), upper part of biozone 2 of Sbaffi et al. (2001).

– ACE-3 (6.4–5.2 cal. ka B.P.). Top: abrupt increase ofHelicosphaera spp. (reaching 3%). Interval: increase ofF. profunda. Equivalence: lower part of ecozone C2of Principato et al. (2003), interval 2 of Giunta et al.(2003), lower part of biozone 1b of Sbaffi et al. (2001).

– ACE-2 (5.2–4.5 cal. ka B.P.). Top: last common occur-rence of B. bigelowii. Interval: high abundance ofHelicosphaera spp., increase of Syracosphaera spp., F.profunda, U. tenuis. Equivalence: upper part of ecozoneC2 of Principato et al. (2003), interval 2 of Giunta et al.(2003), upper part of biozone 1b of Sbaffi et al. (2001).

– ACE-1 (4.5–2.5 cal. ka B.P.). Interval: high abundanceof E. huxleyi (60%) and EHCM (approx. 10%),decrease of Helicosphaera spp. Equivalence: ecozoneC1 of Principato et al. (2003), interval 1 of Giunta et al.(2003), biozone 1a of Sbaffi et al. (2001).

Fig. 8 Comparison between the Aegean coccolithophore ecozones(ACE), the Aegean planktonic foraminifer ecozones (APFE) and thepollen assemblage zones (PAZ), as defined in this paper, and thecorresponding intervals identified by various authors in the easternMediterranean (Principato et al. 2003), Adriatic Sea (Jorissen et al.1993; Capotondi et al. 1999; Giunta et al. 2003) and Tyrrhenian Sea(Jorissen et al. 1993; Sbaffi et al. 2001). Correlation between theschemes of Giunta et al. (2003) and Principato et al. (2003) is after

Principato et al. (2003). Correlation between the schemes ofCapotondi et al. (1999) and Sbaffi et al. (2001) is after Sbaffi et al.(2001). Calibrated ages are based on the age model reconstructed forcore NS-14 core. * Ages from Principato et al. (2003), calibratedusing the CALIB 5.01 program and corrected for a reservoir age of400 years (Siani et al. 2001); ** calibrated age from Principato et al.(2006); *** calibrated ages from Sbaffi et al. (2001)

Geo-Mar Lett (2009) 29:249–267 259

Planktonic foraminifer ecozones

Nine Aegean planktonic foraminifer ecozones (APFE-9 toAPFE-1) are recognised (Fig. 8).

– APFE-9 (~14.5–13.4 cal. ka B.P.). Top: appearance of G.inflata and G. truncatulinoides, high abundance of G.ruber alba (60%), disappearance of N. dutertrei (d).Interval: significant abundance of G. scitula (3%), T.quinqueloba (3%), G. glutinata (10%), N. pachyderma(d) (35%) and N. dutertrei (d) (15%). G. ruber albadecreases in the lower part and increases towards theupper boundary of the interval. Presence of N. pachy-derma (s), increase of N. dutertrei (s) in the upper part.Equivalence: ecozone PF6 of Principato et al. (2003),biozone 5 of Sbaffi et al. (2001, 2004), ecozone 7 ofCapotondi et al. (1999), middle part of zone II of Jorissenet al. (1993).

– APFE-8 (13.4–9.8 cal. ka B.P.). Top: increase of G. ruberalba (45%), appearance of G. ruber rosea. Interval:presence of G. ruber rosea, G. truncatulinoides and G.inflata. Decrease of N. pachyderma, followed by slightincrease in the upper part of the interval (core NS-40).N. dutertrei (s) present in the lower part of this ecozone.Significant presence of G. ruber alba, decreasing in theupper part with a concomitant increase of the cool-waterT. quinqueloba (2%) and G. glutinata (12%). Equiva-lence: ecozone PF5 of Principato et al. (2003), upperpart of biozone 5, and biozones 4 and 3 of Sbaffi et al.(2001, 2004), ecozones 7 (upper part), 6 and 5 ofCapotondi et al. (1999), zone II of Jorissen et al. (1993).

– APFE-7 (9.8–9.1 cal. ka B.P.). Top: temporary disappear-ance of G. inflata, G. sacculifer and N. pachyderma,disappearance of G. truncatulinoides. Interval: highabundance of G. ruber alba (49%) and O. universa(8%), decrease of G. bulloides, gradual increase of G.ruber rosea, high abundance of G. sacculifer (19%),presence of G. inflata, G. truncatulinoides and N.pachyderma (s). Equivalence: ecozone PF4 of Principatoet al. (2003), lower parts of biozone 2 of Sbaffi et al.(2001, 2004) and of ecozone 4 of Capotondi et al. (1999).

– APFE-6 (9.1–8.6 cal. ka B.P.). Top: increase of G.sacculifer (7%), reappearance of G. inflata and N.pachyderma. Interval: significant increase of G. ruberrosea (25%) and O. universa (6%), high abundance ofwarm-water species such as G. rubescens and G. siphon-ifera, high abundance of G. ruber alba (40%) followed bya short-term decrease, significant presence of G. bulloides.Equivalence: lower part of ecozone PF3 of Principato etal. (2003), middle part of biozone 2 of Sbaffi et al. (2001,2004), upper part of ecozone 4 of Capotondi et al. (1999).

– APFE-5 (8.6–7.6 cal. ka B.P.). Top: disappearance of G.inflata and N. pachyderma. Interval: presence of G.

ruber rosea, G. ruber alba, decrease of O. universa,peak abundance of N. pachyderma (s) (1%) and G.inflata (6%). Equivalence: middle parts of ecozone PF3of Principato et al. (2003) and of biozone 2 of Sbaffiet al. (2001, 2004), lower part of ecozone 3 ofCapotondi et al. (1999).

– APFE-4 (7.6–6.4 cal. ka B.P.). Top: temporary disappear-ance of G. sacculifer and reappearance of G. inflata,decrease of G. glutinata. Interval: decrease of G. ruberalba, high abundance of G. ruber rosea (70%) and G.sacculifer (7%). Equivalence: upper parts of ecozone PF3of Principato et al. (2003), of biozone 2 of Sbaffi et al.(2001, 2004) and of ecozone 3 of Capotondi et al. (1999).

– APFE-3 (6.4–5.2 cal. ka B.P.). Top: disappearance of G.inflata. Interval: reappearance and maximum abun-dance of G. inflata (30%), decrease of G. ruber rosea,presence of G. sacculifer and N. pachyderma, signif-icant presence of G. ruber alba, G. rubescens and G.siphonifera in the lower part, immediately followed bya significant decrease. Equivalence: ecozone PF2 ofPrincipato et al. (2003), lower parts of biozone 1b ofSbaffi et al. (2001), biozone 1c of Sbaffi et al. (2004)and of ecozone 2 of Capotondi et al. (1999).

– APFE-2 (5.2–4.2 cal. ka B.P.). Top: increase of G. ruberalba (40%), decrease of G. ruber rosea and O. universa,disappearance of T. quinqueloba. Interval: decrease ofG. ruber alba, high abundance of G. ruber rosea (30%),G. sacculifer (20%) and O. universa (5%), significantpresence of G. bulloides, G. glutinata and T. quinque-loba. Equivalence: lower part of ecozone PF1 ofPrincipato et al. (2003), upper parts of biozone 1b ofSbaffi et al. (2001) and of ecozone 2 of Capotondi et al.(1999) and upper 1c–1b of Sbaffi et al. (2004).

– APFE-1 (4.2–2.5 cal. ka B.P.). Interval: high abundanceof G. ruber alba (23–54%), G. bulloides (29–50%) andG. rubescens, presence of G. sacculifer, G. ruber rosea,O. universa and G. glutinata, small peak of G. inflata(2%). Equivalence: upper part of ecozone PF1 ofPrincipato et al. (2003), biozone 1a of Sbaffi et al.(2001), biozone 1a of Sbaffi et al. (2004), ecozone 1 ofCapotondi et al. (1999), biozones H, G of Geraga et al.(2008). Ecozones APFE 1–7 can be correlated withzone I of Jorissen et al. (1993).

Pollen assemblage zones

In core NS-14, eight pollen assemblage zones (PAZ-1 toPAZ-8) are recognised (Fig. 8).

– PAZ-8 (~12.0–9.9 cal. ka B.P.). High abundance ofArtemisia and Asteraceae, presence of deciduousQuercus and conifers.

260 Geo-Mar Lett (2009) 29:249–267

– PAZ-7 (9.9–9.3 cal. ka B.P.). Maximum abundance ofChenopodiaceae, increase of Artemisia, Centaurea andPoaceae, decrease of most arboreal pollen (Quercusdeciduous and evergreen, Abies, Olea), absence ofPistacia. Low pollen contents.

– PAZ-6 (9.3–8.8 cal. ka B.P.). Increase of Quercus (bothdeciduous and evergreen), Pistacia and Sanquisorba,presence of Abies, decrease of Chenopodiaceae, highpollen contents.

– PAZ-5 (8.8–7.5 cal. ka B.P.). Increase of Chenopodia-ceae, Artemisia, Ephedra and Olea. Decrease ofdeciduous Quercus, Pistacia and Pinus. Absence ofAbies and Picea pollen.

– PAZ-4 (7.5–6.4 cal. ka B.P.). Increase of deciduousQuercus and Pinus. Sharp decrease of Chenopodiaceae,Poaceae, and absence of Artemisia. High pollen contents.

– PAZ-3 (6.4–5.1 cal. ka B.P.). Increase of steppicelements like Chenopodiaceae and Poaceae, and coni-fers like Cedrus. Decrease of Pistacia, deciduousQuercus and Picea in an interval characterised bylower pollen contents.

– PAZ-2 (5.1–4.2 cal. ka B.P.). Increase of deciduousQuercus, Abies, Picea, Pinus, and high pollen contents.Decrease of Chenopodiaceae and Artemisia. Maximumaquatic palynomorph contents.

– PAZ-1 (4.2–2.5 cal. ka B.P.). High abundance ofPoaceae, Chenopodiaceae, Artemisia, Olea, Pistaciaand Sanguisorba. Low values of Quercus and Cedrus.

Discussion and conclusions

In the following, we present a palaeoenvironmentalinterpretation of the findings described above (cf. Table 2).

Late Glacial

~14.5–13.1 cal. ka B.P.

This time span corresponds to the Bølling/Allerød transi-tion, synchronously recognised in the central and easternMediterranean (Sbaffi et al. 2001, 2004; Principato et al.2003), Adriatic (Giunta et al. 2003) and Ionian Sea (Geragaet al. 2008). Plankton ecozones ACE-9 and APFE-9 reflectgenerally warm conditions; the base of the ecozones cannotbe defined, due to the massive flow event recorded in coreNS-14. ACE-9 is characterised by increased abundances ofthe upper–middle photic zone, warm-water (Winter et al.1994) species U. tenuis, confirming the subtropical featuresof this interval. However, the palaeoclimatic curve derivedfrom planktonic foraminifer assemblages indicates climaticfluctuations, similar to those observed in the other parts ofthe eastern Mediterranean, and attributed to the imprint of

Oldest and Older Dryas events (Giunta et al. 2003; Sbaffiet al. 2004). These cool and highly fertile intervals areevidenced in APFE-9 by T. quinqueloba, G. glutinata andG. scitula, species which thrive during the winter period inthe modern eastern Mediterranean (Pujol and VergnaudGrazzini 1995), as well as by the neogloboquadrinids whichconfirm the presence of highly productive surface con-ditions, probably due to increased terrestrial input (Thunell1978). The warmer shifts are reflected mainly by the highabundance of G. ruber alba and, to a lesser extent, ofG. rubescens.

~13.1–9.9 cal. ka B.P.

The increase in abundance of Calciosolenia spp. in ACE-8 indicates a shift to a wetter period (cf. Amore et al. 2004).Cooler conditions are evidenced by the consistently highabundance of E. huxleyi and the increase of EHMC. Thelatter morphotypes are reported to be restricted mainly tocooler intervals (Crudeli et al. 2004), whereas E. huxleyi iscapable of eutrophic adaptation (Young 1994); in particular,it flourishes under high-nutrient conditions and dominatesthe modern Aegean Sea coccolithophore assemblagesduring the winter period (Triantaphyllou et al. 2004;Dimiza et al. 2008). The presence of G. inflata and G.truncatulinoides mainly in the lower part of APFE-8 (~13.0–12.0 cal. ka B.P.) reveals a well-developed deep

Table 2 Late Glacial–Holocene palaeoenvironmental conditions ofthe south-eastern Aegean Sea featured in the present study

Age interval(cal. ka B.P.)

Palaeoecological–palaeoclimatic conditions

~14.5–13.1 Warm period: climatic fluctuations,seasonality

~13.1–9.9 Cool and arid period

9.9–9.3 Increased river runoff, onset of water columnstratification

9.3–9.0 Warm and humid conditions, well-stratified andproductive coastal waters

9.0–8.6 Warm and stratified water column, short coolerintervals

8.6–7.6 Dry and relatively cool period, slight decrease insea surface salinity and stratification, persistentlyhigh productivity in the deep photic zone

7.6–6.9 Humid conditions, warm and stratified watercolumn

6.9–6.4 Wet period with short cooler intervals, decline ofthe deep chlorophyll maximum

6.4–5.2 Decline in temperature and precipitation

5.2–4.2 Very humid and warm conditions

4.2–2.5 Decline in temperature and humidity

Geo-Mar Lett (2009) 29:249–267 261

and cold mixed layer related to cooling (Pujol andVergnaud Grazzini 1995). Analogous conditions have beenrecognised during the same time interval in the Ionian Sea(Geraga et al. 2008). Peaks in abundance of N. pachydermaand N. dutertrei sinistral forms suggest high productivityand cool conditions (Thunell and Sautter 1992). The land-borne imprint of pollen assemblages dominated by Grami-neae, Asteraceae and Artemisia is indicative of dry climaticconditions for the PAZ-8 zone. However, the coexistence ofdeciduous Quercus, Abies and Cedrus implies sufficientprecipitation for the development of forest vegetation infavourable areas, while the sporadic presence of Olea,Pistacia and Sanguisorba after 10.9 cal. ka B.P. is indicativeof frost-free winters. Therefore, we cannot argue for thedetection of severely cold and arid conditions correspondingto the Younger Dryas period during this interval for the easternedge of the Aegean Sea, in contrast to what has been reportedfor other parts of the central and eastern Mediterranean(Geraga et al. 2000, 2008; Sbaffi et al. 2001, 2004).

Holocene

9.9–9.3 cal. ka B.P.

The gradual increase of the nutricline depth proxy (Molfinoand McIntyre 1990) F. profunda in the ACE-7 coccolitho-phore ecozone supports the establishment of stratified con-ditions and the onset of a nutrient-rich environment in thedeep photic zone below the base of S1, consistent with whathas been observed by Principato et al. (2003) west–southwestof Crete. Similar conditions have been recognised for theinterval prior to S1 deposition in the Ionian Sea (Geraga et al.2008); however, they occurred earlier there than in the south-eastern Aegean Sea. In addition to the findings of formerstudies, our data provide evidence for increased freshwaterinput reflected by the gradual increase of Helicosphaera spp.,which indicate a tendency towards lowering of salinity (Floreset al. 1997; Colmenero-Hidalgo et al. 2004) and higherproductivity in the mid photic zone (Ziveri et al. 2004;Crudeli et al. 2006). Lower-salinity conditions and stratifiedwaters are corroborated for the same interval by increasedrelative percentages of the euryhaline species Polysphaeri-dium zoharyi ktana (Wall et al. 1977; Morzadec-Kerfourn1988; Sanchez Goni et al. 1999; Marret and Zonneveld 2003).Planktonic foraminifers in the APFE-7 ecozone support theabovementioned findings. The assemblages are characterisedby the appearance of G. sacculifer and the increasedabundance of N. pachyderma sinistral forms, which indicatethe occurrence of a shallow pycnocline (Fenton et al. 2000),possibly related to periods of strong river runoff (Rossignol-Strick 1985) and associated with the development of a deepchlorophyll maximum in the lower photic zone. G. bulloidesis also indicative of high primary productivity; G. ruber alba,

a warm oligotrophic surface dweller (Thunell 1978; Pujol andVergnaud Grazzini 1995), shows a strong increasing trendindicative of the onset of water column stratification.Lowering of sea surface temperatures, indicated by thepalaeoclimatic curves at the base of APFE-7, is consistentwith the increase of cool-water indicators (T. quinqueloba, G.inflata, G. glutinata) immediately before the deposition of S1.In addition, lower sea surface temperatures are suggested forthe interval preceding S1a, by the high relative abundances ofN. labyrinthus and P. dalei (e.g. Combourieu-Nebout et al.1999; Sanchez Goni et al. 1999; Sangiorgi et al. 2003). Theabundances of Cedrus, Quercus and other deciduous taxafeaturing in the PAZ-7 zone are indicative of the presence ofsome conifer and open deciduous forests in favourable landareas with sufficient moisture. However, the recorded delay ofwoodland expansion during the early Holocene (relative tohigher latitudes) reveals a difference in synchronicity betweenthe plankton ecozones ACE-7, APFE-7 and the pollenassemblage zone PAZ-7. This delay has also been observedin onshore deposits from SW Turkey (Eastwood et al. 1999)and has been correlated with climatic aridity (Roberts andWright 1993). In the present study, the curve of Chenopodia-ceae, typical for very cold and dry periods (e.g. Rossignol-Strick 1995), shows a pronounced maximum. It is ofparticular interest that the expansion of Chenopodiaceaeduring this time interval is not associated with the phasedescribed by Rossignol-Strick (1995, 1999) correlating withthe Younger Dryas chronozone, but possibly reflects theexistence of a salt marshy coastal zone (e.g. Geraga et al.2000). Consistent with these findings, Van Zeist et al. (1975)and Vermoere et al. (1999) attributed the increased dryvegetation elements observed around 9.0 uncal. ka B.P. insouth-western Turkey to higher temperatures during summerintervals or longer periods of summer drought.

9.3–9.0 cal. ka B.P. (S1a deposition)

The increased values of Helicosphaera spp., followedby the high abundance of tropical–subtropical F. profundain the lower part of S1a, confirm the development of anutrient-rich environment in the middle photic zone andfurther support the lowering of salinity for the ACE-6becozone. A shallow pycnocline and strong runoff aresuggested by the high abundance of G. sacculifer in theAPFE-6 ecozone, as well as by the high aquatic palyno-morph contents in the PAZ-6 zone. Previous studies (e.g.Aksu et al. 1995) have suggested that S1 deposition in theAegean Sea coincided with a decrease of 1–1.5‰ in seasurface salinity. The shift of the palaeoclimatic curvetowards increased values is the main feature of the APFE-6 ecozone, caused by the high abundance of warm-waterspecies (G. ruber alba and O. universa); G. ruber rosea isfurther indicative of warmer conditions (Hemleben et al.

262 Geo-Mar Lett (2009) 29:249–267

1989), associated with enhanced water column stratifica-tion. Alkenone data from the northern Aegean (Gogou et al.2007a) indicate that sea surface temperatures increased to~21°C during the deposition of the lower half of the S1sapropel. Maximum relative percentages of heterotrophicdinoflagellate cysts and high values of the H/A index in thelower part of the S1a sapropel (around 9.2 cal. ka B.P.) arealso consistent with elevated nutrient levels (Zonneveld etal. 2001, 2007). Moreover, pollen assemblages reflecting aterrestrial signature in PAZ-6 show the increase of bothdeciduous and evergreen taxa such as Quercus, Pistaciaand Olea at the base of S1a, indicating the onset offorestation and the development of Mediterranean vegeta-tion belts as a result of warm and humid conditions duringthis time period.

9.0–8.6 cal. ka B.P. (S1a deposition)

Former workers have interpreted the S1a depositionalinterval as a generally warm period (e.g. Geraga et al.2008); however, the present results demonstrate short-termpulses of cooler climate. Thus, this period is more complexthan previously thought, supporting the establishment of amore detailed ecostratigraphy compared to the existingschemes (see Fig. 8). In particular, warm conditions withcolder shifts are suggested by the palaeoclimatic curve andthe faunal assemblages which prevail in the APFE-6ecozone in the upper part of S1a. Strong warming ofsurface waters and water column stratification can beinferred by the high values of G. ruber rosea, G. rubescens,G. siphonifera and the dinocyst L. machaerophorum.Indications of cooler shifts are supported in ACE-6a bythe gradual increase of Umbilicosphaera spp., a groupcapable of eutrophic adaptation (Young 1994), this beingconsistent with nutrient redistribution in surface watersreported by Principato et al. (2006). In addition, the higherabundance of the dinoflagellate P. reticulata might beassociated with eutrophic surface waters (Marret andZonneveld 2003), while the considerable abundance of I.aculeatum implies not only warmer sea surface temper-atures but also the existence of eutrophic environments inwhich it can constitute up to 20% of the cyst association(Marret and Zonneveld 2003).

8.6–7.6 cal. ka B.P. (S1 interruption)

Our high sedimentation rate records enable the detaileddocumentation of palaeoecological conditions within the S1interruption interval, which lasted longer in the south-eastern Aegean than in the Ionian Sea (8.5–8.0 cal. ka B.P.;Geraga et al. 2008). Higher productivity in the deep photiczone (cf. high F. profunda levels, decreasing only in thelower part of the interruption in the ACE-5 ecozone) co-

occurs with production in a relatively less saline surfacelayer (cf. moderate abundance of Helicosphaera spp., andsporadic presence of B. bigelowii, an indicator of lowsalinity; Negri and Giunta 2001; Giunta et al. 2003). Aslight breakdown of stratification can be suggested by thedecrease of G. ruber rosea, the dramatic reduction of G.rubescens, O. universa and the dinocyst L. machaeropho-rum, and the presence of G. glutinata and N. pachydermadextral specimens in the APFE-5 ecozone. Terrestrialenvironmental response to climate variability is evidencedby a small expansion of steppe vegetation and a concurrentdecline in runoff proxies recorded in the PAZ-5 zone,pointing to a drier period. The presence of Ephedra mayreflect a decrease in temperature (Mudie et al. 2002). Inaddition, the absence of Pistacia for short intervals duringthe S1 interruption could represent short periods oftemperature decrease (Rossignol-Strick 1995).

7.6–6.9 cal. ka B.P. (S1b deposition)

Within the lower part of S1b, a small but clear peak of B.bigelowii, coupled with relatively high abundances ofHelicosphaera spp. and F. profunda, suggest the persistentpresence of less saline surface waters and stratifiedconditions for the ACE-4b coccolithophore ecozone. Theincrease in abundance of G. ruber rosea, O. universa andG. rubescens in the APFE-4 ecozone represents furtherevidence of water column stratification. Stratification andlower-salinity conditions are supported by highest relativeabundance of the autotroph euryhaline dinoflagellate L.machaerophorum (Wall et al. 1977; Morzadec-Kerfourn1988; Sanchez Goni et al. 1999; Marret and Zonneveld2003). The expansion of mixed deciduous forests on land(PAZ-4 zone) and the increase in aquatic palynomorphcontents indicate a humid period.

6.9–6.4 cal. ka B.P. (S1b deposition)

F. profunda displays a distinct reduction in the coccolitho-phore ecozone ACE-4a, revealing a decline in productivity inthe deep photic zone within the upper part of S1b. The samepattern is recorded within the upper APFE-4, associated withthe decrease of warm foraminiferal species (e.g. G. ruberrosea, O. universa and G. rubescens) which precedes thecessation of sapropelic conditions. Simultaneously, there isan abrupt increase of the hyposaline and relatively eutrophicnannofossil indicator B. bigelowii (Boalch 1987; Negri andGiunta 2001). A similar increase of this species has beenrecorded for the same time interval in the eastern Ionian(Negri and Giunta 2001), interpreted as enhanced stratifica-tion, and in the area west–southwest of Crete (Principatoet al. 2003), related to runoff conditions during a wet andcold period. Our combined plankton data favour the latter

Geo-Mar Lett (2009) 29:249–267 263

hypothesis, suggesting climate cooling before the end of S1sapropelic conditions. Obviously, sapropel S1 lasted longerin the south-eastern Aegean—and also at other easternMediterranean locations (Negri and Giunta 2001; Principatoet al. 2003)—than in the Ionian Sea, where Geraga et al.(2008) report an age of 7.0 cal. ka B.P. for the top of S1associated with a cold event. A similar cooling eventbetween ~6.8–6 cal. ka B.P. has been identified in thesouthern Adriatic Sea (Sangiorgi et al. 2003) and between8–6.5 cal. ka B.P. in the South Aegean (Geraga et al. 2005;stadial C69-ST2). Interestingly, Gogou et al. (2007a)reported pronounced sea surface temperature coolingsthroughout the deposition of S1 and, in particular, duringits later part, at least in the North Aegean. The palaeoclimaticcurves obtained in this study support this climatic instabilityand the presence of cooling within the upper S1b, bound tothe strengthening of northerly winds which affected subtrop-ical/tropical control of the regional hydrography and ecosys-tems of the south-eastern Aegean Sea (e.g. Casford et al.2002; Rohling et al. 2002b).

6.4–5.2 cal. ka B.P.

The foraminiferal patterns recorded during this time intervalimply openmarine conditions and pronounced vertical mixingof the water column, as indicated by the peak abundance ofG.inflata in the APFE-3 zone and the associated lowering ofsea surface temperature recorded in the palaeoclimaticcurves. Higher productivity in the surface layers is indicatedby the increasing trend of E. huxleyi in the ACE-3coccolithophore ecozone. This temperature decline is sup-ported by the expansion of coniferous forests recorded in thePAZ-3 zone, while expansion of steppic elements in theneighbouring land areas suggests a decline in precipitation.

5.2–4.2 cal. ka B.P.

A distinct warming phase has been interestingly recognisedonly in the shallower NS-14 record, indicated by theforaminiferal palaeoclimatic curve. It is evidenced in theAPFE-2 ecozone by the significant increase of G. ruberrosea and O. universa, and the temporary disappearance ofG. inflata. Moreover, intense freshwater input is docu-mented by the high abundance of the foraminifer G.sacculifer and the occurrence of T. quinqueloba, explainedby the tolerance of this species to fairly low salinity and/orenhanced fertility in surface waters (Rohling et al. 1997).Additional evidence for higher river runoff is provided bythe increase of the coccolithophores Helicosphaera spp.and Syracosphaera spp., both indicators of lower salinity(Flores et al. 1997; Colmenero-Hidalgo et al. 2004),followed by high abundance of Calciosolenia spp. andminor peaks of B. bigelowii in the ACE-2 ecozone. The

establishment of stratified conditions in the water column ismarked by the prominent peak of F. profunda. On land, theexpansion of mixed deciduous forest, together with themaxima recorded for aquatic palynomorphs, indicatesstrong precipitation for the PAZ-2 zone. The presence ofPicea in the area is indicative of north wind outbursts(Mudie et al. 2002).

4.2–2.5 cal. ka B.P.

A cool interval, bearing the features of present-day easternMediterranean winter plankton assemblages, is recognisedwithin coccolithophore ACE-1, based on the increase of E.huxleyi and of cool EHMC morphotypes (cf. Crudeli et al.2004). High primary productivity in a cold mixed layer isrecorded by the occurrence of G. bulloides, G. inflata, G.truncatulinoides and heterotrophic dinocysts, whereas thepresence of G. glutinata points to cooling events for theAPFE-1 ecozone. This event is also recorded in the DeadSea at ~3.5 ka (Migowski et al. 2006) and in the southernAdriatic Sea (Sangiorgi et al. 2003), implying spring andpossibly summer temperature decrease. PAZ-1 reflects aterrestrial environment with open vegetation cover andindications of human disturbance to vegetation. The peak ofPistacia observed at ~3 cal. ka B.P. suggests a notable rise inwinter temperatures (Rossignol-Strick 1995; Combourieu-Nebout et al. 1998; Giunta et al. 2003), and is consistentwith the increased abundance of the warm-water dinofla-gellate species L. machaerophorum and O. israelianumaround 3.2 cal. ka B.P.

In conclusion, assessments of plankton and pollenabundance patterns, together with palaeoclimatic curvesinferred from planktonic foraminiferal assemblages in bothshallow (NS-14) and deeper (NS-40) core records, havesupported the establishment of a detailed ecostratigraphicscheme spanning the last ~14.5 cal. ka at the south-easternmargin of the Aegean Sea. The Aegean ecozones providedby this work are well correlated with all existing easternMediterranean ecozonal schemes. Nevertheless, the highsedimentation rates which characterise the study area haveenabled the detection of even minor and brief climatic eventsnot recognised to date, probably linked to high-latitudeHolocene climate influence superimposed on eastern Med-iterranean Sea hydrography and ecosystem dynamics duringsubtropical/tropical S1 deposition times.

Acknowledgements We acknowledge financial support from thePythagoras I project (EU and Greek Ministry of Education, EPEAEKII) and partial support from the European Science Foundation (ESF)under the EUROCORES Programme EuroCLIMATE, through con-tract no. ERAS-CT-2003-980409 of the European Commission, DGResearch, FP6. Critical comments by two anonymous referees and thejournal editors have proved essential in improving the manuscript.

264 Geo-Mar Lett (2009) 29:249–267

References

Aksu AE, Yaşar D, Mudie PJ, Gillespie H (1995) Late Glacial-Holocene paleoclimatic and paleoceanographic evolution of theAegean Sea: micropaleontological and stable isotopic evidence.Mar Micropaleontol 25:1–28

Amore FO, Gaffau M, Massa B, Morabito S (2004) Late Pleistocene-Holocene paleoclimate and related paleoenvironmental changesas recorded by calcareous nannofossils and planktonic forami-nifera assemblages in the southern Tyrrhenian Sea (CapePalinuro, Italy). Mar Micropaleontol 52:255–276

Anastasakis G, Stanley DJ (1986) Uppermost sapropel, EasternMediterranean: paleoceanography and stagnation. Natl GeogrRes 2:179–197

Boalch GT (1987) Changes in the phytoplankton of the westernEnglish Channel in recent years. Br Phycol J 22:225–235

Capotondi L, Borsetti A, Morigi C (1999) Foraminiferal ecozones, ahigh resolution proxy for the late Quaternary biochronology inthe central Mediterranean Sea. Mar Geol 153:253–274

Capotondi L, Soroldoni E, Principato MS, Corselli C (2004) LateQuaternary foraminiferal distributions: problems related to sizefraction. In: Coccioni R, Galeotti S, Lirer F (eds) Proc 1st ItalianMeet Environmental Micropaleontology. Grzybowski FoundationSpec Publ 9:1–6

Capotondi L, Principato MS, Morigi C, Sangiorgi F, Maffioli P,Giunta S, Negri A, Corselli C (2006) Foraminiferal variationsand stratigraphic implications to the deposition of sapropel S5 inthe eastern Mediterranean. Palaeogeogr Palaeoclimatol Palae-oecol 235:48–65

Casford JSL, Rohling EJ, Abu-Zied R, Cooke S, Fontanier C, Leng M,Lykousis V (2002) Circulation changes and nutrient concentrationsin the late Quaternary Aegean Sea: a nonsteady state concept forsapropel formation. Paleoceanography 17:1024–1034

Colmenero-Hidalgo E, Flores JA, Sierro FJ, Bárcena MÁ, Löwemark L,Schönfeld J, Grimalt JO (2004) Ocean surface water response toshort-term climate changes revealed by coccolithophores from theGulf of Cadiz (NE Atlantic) and Alboran Sea (W Mediterranean).Palaeogeogr Palaeoclimatol Palaeoecol 205:317–336

Combourieu-Nebout N, Paterne M, Turon JL, Siani G (1998) A highresolution record of the last deglaciation in central MediterraneanSea: palaeovegetation and palaeohydrological evolution. QuatSci Rev 17:303–317

Combourieu-Nebout N, Londeix L, Baudin F, Turon JL, von GrafensteinR, Zahn R (1999) Quaternary marine and continental paleoenviron-ments in the western Mediterranean (Site 976, Alboran Sea):palynological evidence. In: Zahn R, Comas MC, Klaus A (eds)Proc Ocean Drilling Program Sci Results 161:457–468

Crudeli D, Young JR, Erba E, de Lange GJ, Henriksen K, Kinkel H,Slomp CP, Ziveri P (2004) Abnormal carbonate diagenesis inHolocene–late Pleistocene sapropel-associated sediments fromthe Eastern Mediterranean; evidence from Emiliania huxleyicoccolith morphology. Mar Micropaleontol 52:217–240

Crudeli D, Young JR, Erba E, Geisen M, Ziveri P, de Lange GJ, Slomp CP(2006) Fossil record of holococcoliths and selected hetero-holococcolith associations from the Mediterranean (Holocene–latePleistocene): evaluation of carbonate diagenesis and palaeoecolog-ical–palaeocenographic implications. Palaeogeogr PalaeoclimatolPalaeoecol 237:191–212

De Rijk S, Hayes A, Rohling EJ (1999) Eastern Mediterraneansapropel S1 interruption: an expression of the onset of climaticdeterioration around 7 ka BP. Mar Geol 153:337–343

Dimiza M, Triantaphyllou MV, Dermitzakis MD (2008) Seasonalityand ecology of living coccolithophores in E. Mediterraneancoastal environments (Andros Island, Middle Aegean Sea).Micropaleontology 54(2):159–175

Eastwood WJ, Roberts N, Lamb HF, Tibby JC (1999) Holoceneenvironmental catchment in southwest Turkey: a palaeoecolog-ical record of lake and catchment-related changes. Quat Sci Rev18:671–695

Emeis KC, Struck U, Schulz HM, Rosenberg R, Bernasconi S,Erlenkeuser H, Sakamoto T, Martinez-Ruiz F (2000) Tempera-ture and salinity variations of Mediterranean Sea surface watersover the last 16, 000 years from records of planktonic stableoxygen isotopes and alkenone unsaturation ratios. PalaeogeogrPalaeoclimatol Palaeoecol 158:259–280

Emeis KC, Schulz H, Struck U, Rossignol-Strick M, Erlenkeuser H,Howell MW, Kroon D, Mackensen A, Ishizuka S, Oba T,Sakamoto T, Koizumi I (2003) Eastern Mediterranean surfacewater temperatures and δ18O composition during deposition ofsapropels in the late Quaternary. Paleoceanography 18(1):1005.doi:10.1029/2000PA00061

Facorellis Y, Maniatis Y, Kromer B (1998) Apparent 14C ages of marinemollusk shells from a Greek island: calculation of the marinereservoir effect in the Aegean Sea. Radiocarbon 40:963–973

Fenton M, Geiselhart S, Rohling EJ, Hemleben C (2000) Planktoniczones in the Red Sea. Mar Micropaleontol 40:277–294

Flemming BW, Delafontaine MT (2000) Physical properties of muddyintertidal sediments: some applications, misapplications and non-applications. Cont Shelf Res 20:1179–1197

Flores JA, Sierro FJ, Francés G, Vázquez A, Zamarreňo I (1997) Thelast 100, 000 years in the western Mediterranean: sea surfacewater and frontal dynamics as revealed by coccolithophores. MarMicropaleontol 29:351–366

Geraga M, Tsaila-Monopolis S, Ioakeim C, Papatheodorou G,Ferentinos G (2000) Evaluation of palaeoenvironmental changesduring the last 18, 000 years in the Myrtoon basin, SW AegeanSea. Palaeogeogr Palaeoclimatol Palaeoecol 156:1–17

Geraga M, Tsaila-Monopolis S, Ioakim C, Papatheodorou G,Ferentinos G (2005) Short-term climate changes in the southernAegean Sea over the last 48,000 years. Palaeogeogr Palae-oclimatol Palaeoecol 220:311–332

Geraga M, Mylona G, Tsaila-Monopoli S, Papatheodorou G,Ferentinos G (2008) Northeastern Ionian Sea: palaeoceano-graphic variability over the last 22 ka. J Mar Syst 74:623–638

Giunta S, Negri A, Morigi C, Capotondi L, Combourieu-Nebout N,Emeis KC, Sangiorgi F, Vigliotti L (2003) Coccolithophoridecostratigraphy and multi-proxy paleoceanographic reconstruc-tion in the Southern Adriatic Sea during the last deglacial time(Core AD91–17). Palaeogeogr Palaeoclimatol Palaeoecol190:39–59

Gogou A, Bouloubassi I, Stephanou EG (2000) Marine organicgeochemistry of the Eastern Mediterranean: 1. Aliphatic andpolyaromatic hydrocarbons in Cretan Sea surficial sediments.Mar Chem 68:265–282

Gogou A, Bouloubassi I, Lykousis V, Arnaboldi M, Gaitani P, Meyers PA(2007a) Organic geochemical evidence of abrupt late glacial-Holocene climate changes in the North Aegean Sea. PalaeogeogrPalaeoclimatol Palaeoecol 256:1–20

Gogou A, Katsouras G, Emeis K, Bouloubassi I, Lykousis V,Triantaphyllou M, Arnaboldi M, Meyers PA (2007b) The organicgeochemical evidence of Late Glacial-Holocene climate instabilityin the Aegean Sea: teleconnections with North Atlantic paleoclimatevariability. AGU 88 (52) Fall meeting Abstract PP43B-1280

Hemleben C, Spindler M, Anderson OR (1989) Modern planktonicForaminifera. Springer, Berlin Heidelberg New York

Ignatiades L, Psarra S, Zervakis V, Pagou K, Souvermezoglou E,Assimakopoulou G, Gotsis-Skretas O (2002) Phytoplankton size-based dynamics in the Aegean Sea (Eastern Mediterranean). JMar Syst 36:11–28

Jorissen FJ, Asioli A, Borsetti AM, Capotondi L, De Visser JP, Hilgen FJ,Rohling EJ, Van der Borg K, Vergnaud-Grazzini C, Zachariasse WJ

Geo-Mar Lett (2009) 29:249–267 265

(1993) Late Quaternary central Mediterranean biochronology. MarMicropaleontol 21:169–189

Lacombe H, Tchernia P (1972) Caractères hydrolologiques et circulationdes eaux en Méditerranée. In: Stanley DJ (ed) The MediterraneanSea. Dowden Hutchinson & Ross, Stroudsburg, PA, pp 26–36

Lourens LJ, Hilgen FJ, Gudjonsson L, Zachariasse WJ (1992) LatePliocene to Early Pleistocene astronomically forced sea surfaceproductivity and temperature variations in the Mediterranean.Mar Micropaleontol 19:49–78

Lykousis V, Chronis G, Tselepides A, Price NB, Theocharis A, Siokou-Fragou I, Wambeke F, Danovaro R, Stavrakakis S, Duineveld G,Georgopoulos D, Ignatiades L, Souvermezoglou A, Voutsinou-Taliadouri F (2002) Major outputs of the recent multidisciplinarybiogeochemical researches undertaken in the Aegean Sea. J MarSyst 33(34):313–334

Marino G (2007) Paleoceanography of the interglacial easternMediterranean Sea. LPP Foundation Contr Ser 24, Utrecht

Marret F, Zonneveld K (2003) Atlas of modern organic-walleddinoflagellate cyst distribution. Rev Palaeobot Palynol 125:1–200

Mercone D, Thomson J, Abu-Zied RH, Croudace IW, Rohling EJ(2001) High-resolution geochemical and micropaleontologicalprofiling of the most recent eastern Mediterranean sapropel. MarGeol 177:25–44

Migowski C, Mordechai S, Prasad S, Negendank JFW, Agnon A(2006) Holocene climate variability and cultural evolution in theNear East from the Dead Sea sedimentary record. Quat Res66:421–431

Molfino B, McIntyre A (1990) Precessional forcing of nutriclinedynamics in the Equatorial Atlantic. Science 249:766–769

Moodley L, Middelburg JJ, Herman PMJ, Soetaert K, de Lange GJ(2005) Oxygenation and organic-matter preservation in marinesediments: direct experimental evidence from ancient organiccarbon-rich deposits. Geology 33(11):889–892

Moore PD, Webb JA, Collinson ME (1991) Pollen analysis. Black-well, Oxford

Morzadec-Kerfourn MT (1988) Palaeoclimates and paleoenviron-ments from the Lateglacial to recent, in the Eastern Mediterra-nean, east of the Nile delta, the contribution of organic-walledmicrofossils. Bull Centres de Recherches Exploration-ProductionElf Aquitaine 12:267–275

Mudie PJ, Rochon A, Aksu AE (2002) Pollen stratigraphy of LateQuaternary cores from Marmara Sea: land-sea correlation andpaleoclimatic history. Mar Geol 190:233–260

Negri A, Giunta S (2001) Calcareous nannofossil paleoecology in thesapropel S1 of the eastern Ionian Sea: paleoceanographic implica-tions. Palaeogeogr Palaeoclimatol Palaeoecol 169:101–112

Papanikolaou DJ, Nomikou P (2001) Tectonic structure and volcaniccentres at the eastern edge of the Aegean volcanic arc aroundNisyros Island. Bull Geol Soc Greece 34(1):289–296

Perissoratis C, Piper DJW (1992) Age, regional variation andshallowest occurrence of S1 sapropel in the northern AegeanSea. Geo-Mar Lett 12:49–53. doi:10.1007/BF02092108

Poulos SE, Drakopoulos PG, Collins MB (1997) Seasonal variabilityin sea surface oceanographic conditions in the Aegean Sea(Eastern Mediterranean): an overview. J Mar Syst 13:225–244

Principato M, Giunta S, Corselli C, Negri A (2003) Late Pleistocene–Holocene planktonic assemblages in three box-cores from theMediterranean Ridge area (west-southwest of Crete): paleoeco-logical and palaeoceanographic reconstruction of sapropel S1interval. Palaeogeogr Palaeoclimatol Palaeoecol 190:61–77

Principato M, Crudeli D, Ziveri P, Slomp CP, Corselli C, Erba E, deLange GJ (2006) Phyto- and zooplankton paleofluxes during thedeposition of sapropel S1 (eastern Mediterranean): biogeniccarbonate preservation and paleoecological implications. Palae-ogeogr Palaeoclimatol Palaeoecol 235:8–27

Pujol C, Vergnaud Grazzini C (1995) Distribution patterns of liveplanktic foraminifers as related to regional hydrography andproductive systems of the Mediterranean Sea. Mar Micropaleon-tol 25:187–217

Ramsey CB, Manning SW, Galimberti M (2004) Dating the volcaniceruption at Thera. Radiocarbon 46:325–344

Reille M (1992) Pollen et spores d'Europe et d'Afrique du Nord.Laboratoire de Botanique Historique et Palynologie URA CNRS,Marseille

Roberts N, Wright HE (1993) Vegetation, lake level and climatichistory of the Near East and southern Asia. In: Wright HE,Kutzbach JE, Webb T, RuddimanWF, Street-Perrott FA, Bartlein PJ(eds) Global climates since the last glacial maximum. University ofMinnesota Press, Minneapolis, MN, pp 409–429

Rohling EJ, Jorissen FJ, Vergnaud Grazzini C, Zachariasse WJ (1993)Northern Levantine and Adriatic quaternary planktic foraminif-era: reconstruction of paleoenvironmental gradients. Mar Micro-paleontol 21:191–218

Rohling EJ, Jorissen FJ, De Stigter HC (1997) 200 Year interruptionof Holocene sapropel formation in the Adriatic sea. J Micro-paleontol 16:97–108

Rohling EJ, Cane TR, Cooke S, Sprovieri M, Bouloubassi I, Emeis KC,Schiebel R, Kroon D, Jorissen FJ, Lorre A, Kemp AES (2002a)African monsoon variability during the previous interglacialmaximum. Earth Planet Sci Lett 202:61–75

Rohling EJ, Mayewski PA, Abu-Zied RH, Casford JSL, Hayes A(2002b) Holocene atmosphere–ocean interactions: records fromGreenland and the Aegean Sea. Climate Dynamics 18:587–593

Rohling EJ, Hopmans EC, Sinninghe Damsté JS (2006) Water columndynamics during the last interglacial anoxic event in theMediterranean (sapropel S5). Paleoceanography 21:1–8

Rossignol-Strick M (1985) Mediterranean Quaternary sapropels, animmediate response of the African monsoon to variation ofinsolation. Palaeogeogr Palaeoclimatol Palaeoecol 49:237–263

Rossignol-Strick M (1995) Sea-land correlation of pollen records inthe Eastern Mediterranean for the glacial-interglacial transition:biostratigraphy versus radiometric time-scale. Quat Sci Rev14:893–915

Rossignol-Strick M (1999) The Holocene climatic optimum andpollen records of sapropel 1 in the eastern Mediterranean, 9000–6000 BP. Quat Sci Rev 18:515–530

Sanchez Goni MF, Eynaud F, Turon JL, Shackleton NJ (1999) Highresolution palynological record off the Iberian margin: directland–sea correlation for the Last Interglacial complex. EarthPlanet Sci Lett 171:123–137

Sangiorgi F, Capotondi L, Combourieu Nebout N, Vigliotti L,Brinkhuis H, Giunta S, Lotter AF, Morigi C, Negri A, Reichart G-J(2003) Holocene seasonal sea-surface temperature variation in thesouthern Adriatic Sea inferred from amultiproxy approach. J Quat Sci18:723–732

Sbaffi L, Wezel FC, Kallel N, Paterne M, Cacho I, Ziveri P,Shackleton N (2001) Response of the pelagic environment topaleoclimatic changes in the central Mediterranean Sea duringthe Late Quaternary. Mar Geol 178:39–62

Sbaffi L, Wezel FC, Curzi G, Zoppi U (2004) Millennial- tocentennial-scale palaeoclimatic variations during Termination Iand the Holocene in the central Mediterranean Sea. Global PlanetChange 40:201–217

Siani G, Paterne M, Michel E, Sulpizio R, Sbrana A, Arnols M,Haddad G (2001) Mediterranean sea surface radiocarbon reser-voir age changes since the Last Glacial Maximum. Science294:1917–1919

Targarona J (1997) Climatic and oceanographic evolution of theMediterranean region over the last glacial-interglacial transition.LPP Contr Ser 7, Utrecht

266 Geo-Mar Lett (2009) 29:249–267

Theocharis Α, Georgopoulos D (1993) Dense water formation overthe Samothraki and Limnos Plateaux in the North Aegean Sea(Eastern Mediterranean Sea). Cont Shelf Res 13(8/9):919–939

Theocharis A, Nittis H, Kontoyiannis E, Papageorgiou E, Balopoulos E(1999) Climatic changes in the Aegean Sea influence the EasternMediterranean thermohaline circulation (1986–1997). GeophysRes Lett 26(11):1617–1620

Thunell RC (1978) Distribution of recent of planktonic foraminifera insurface sediments of the Mediterranean Sea. Mar Micropaleontol3:147–173

Thunell R, Sautter LR (1992) Planktonic foraminiferal faunal andstable isotope indices of upwelling: a sediment trap study in theSan Pedro Basin, South California Bight. In: Summerhayes CP,Prell WL, Emeis KC (eds) Upwelling systems: evolution sincethe Early Miocene. Geol Soc Lond Spec Publ 64:77–91

Triantaphyllou MV, Ziveri P, Tselepides A (2004) Coccolithophoreexport production and response to seasonal surface watervariability in the oligotrophic Cretan Sea (ΝΕ Mediterranean).Micropaleontology 50:127–144

Triantaphyllou MV, Gogou A, Lykousis V, Bouloubassi I, Ziveri P,Rosell-Mele A, Kouli K, Dimiza M, Papanikolaou M, Gaitani P,Katsouras G, Dermitzakis MD (2006) Primary production trendsand response of terrestrial environments in SE Aegean core NS-14. A multiproxy approach. Geophys Res Abstr 8:08347

Triantaphyllou M, Antonarakou A, Kontakiotis G, Dimiza M, Ziveri P,Mortyn G, Lianou V, Lykousis V, Dermitzakis MD (2007)Calcareous nannofossil and planktonic foraminiferal assemblagesand paleoecological reconstruction of sapropel S1 in SE AegeanSea. European Geosciences Union, 15-20 April 2007, Vienna,Geophys Res Abstr 9:07805

van Santvoort PJM, de Lange GJ, Langereis CG, Dekkers MJ, Paterne M(1997) Geochemical and paleomagnetic evidence for the occurrenceof "missing" sapropels in eastern Mediterranean sediments. Paleo-ceanography 12:773–786

Van Zeist W, Wordring H, Stapert D (1975) Late Quaternary vegetationand climate of southwestern Turkey. Palaeohistoria 17:53–143

Vermoere M, Degryse P, Vanhecke L, Muchez P, Paulissen E, Smets E,Waelkens M (1999) Pollen analysis of two travertine sections inBasköy (southwestern Turkey): implications for environmentalconditions during the early Holocene. Rev Palaeobot Palynol105:93–110

Wall D, Dale B, Lohmann GP, Smith WK (1977) The environmentaland climatic distribution of dinoflagellate cysts in modern marinesediments from regions in the north and south Atlantic Oceansand adjacent seas. Mar Micropaleontol 2:121–200

Winter A, Jordan RW, Roth P (1994) Biogeography of livingcoccolithophores in oceanic waters. In: Winter A, Siesser WG(eds) Coccolithophores. Cambridge University Press, Cambridge,pp 13–27

Young JR (1994) Functions of coccoliths. In: Winter A, Siesser WG(eds) Coccolithophores. Cambridge University Press, Cambridge,pp 63–82

Zervakis V, Georgopoulos D, Drakopoulos PG (2000) The role of theNorth Aegean in triggering the recent Eastern Mediterraneanclimatic changes. J Geophys Res 105:26103–26116

Ziveri P, Baumann KH, Böckel B, Bollmann J, Young J (2004)Biogeography of selected coccolithophores in the Atlantic Ocean,from Holocene sediments. In: Thierstein H, Young J (eds)Coccolithophores: from molecular processes to global impact.Springer, Berlin Heidelberg New York, pp 403–428

Zonneveld KAF, Versteegh GJM, de Lange GJ (2001) Palaeoproduc-tivity and post-depositional aerobic organic matter decayreflected by dinoflagellate cyst assemblages of the EasternMediterranean S1 sapropel. Mar Geol 172:181–195

Zonneveld KAF, Bockelmann F, Holzwarth U (2007) Selectivepreservation of organic-walled dinoflagellate cysts as a tool toquantify past net primary production and bottom water oxygenconcentrations. Mar Geol 237:109–126

Geo-Mar Lett (2009) 29:249–267 267