Palynology of sapropelic layers from the Marmara Sea

Palynology of sapropelic layers from the Marmara Sea

Marine Geology 190 (2002) 35^46 www.elsevier.com/locate/margeo Palynology of sapropelic layers from the Marmara Sea Hu«lya Caner  , Oya Algan Instit...

427KB Sizes 0 Downloads 40 Views

Marine Geology 190 (2002) 35^46 www.elsevier.com/locate/margeo

Palynology of sapropelic layers from the Marmara Sea Hu«lya Caner  , Oya Algan Institute of Marine Sciences and Management, Istanbul University, Vefa 34470, Turkey Received 28 February 2001; accepted 19 February 2002

Abstract Palynological records of the sediments from the deep basins of the Marmara Sea revealed four palynological zones indicating the changing climatic conditions during Late Glacial to Holocene. The lower two zones were defined by the high abundance of Artemisia and Chenopodiaceae that suggest the climate of the source area was cold and arid. However, the establishment of Mediterranean warm and wet climate and corresponding regression of continental aridity were shown by the decreasing Chenopodiaceae and increasing pollen grains of humid-type vegetation in Zone C. The highest total pollen with the dominance of Quercus identifies the deglaciation and coincides with the formation of sapropelic layers in the Marmara Sea. The warmer and humid condition of deglaciation was indicated with diversified moisture-demanding deciduous and coniferous pollen grains in Zone B. These pollen assemblages and their distribution pattern indicate that the source area was warm and wet, during the deposition of sapropelic layers in the Marmara Sea, and also reflect signals of Black Sea origin. 0 2002 Elsevier Science B.V. All rights reserved. Keywords: Marmara Sea; palynology; sapropelic layers

1. Introduction Palynological investigations of the Late Quaternary sedimentary sequences have been successfully used in constructing the paleoclimatic^paleoceanographic conditions in association with the sapropel formations in the Mediterranean Sea (Cheddadi et al., 1991; Aksu et al., 1995, 1999; Rossignol-Strick and Paterne, 1999; Grega et al., 2000). The formation of sapropel/sapropelic sediments is associated with oxygen de¢ciency/anoxia conditions at the bottom and coupled with high

* Corresponding author.

amount of organic matter. Such conditions in marine environment favor the preservation of continuous pollen records. The pollen assemblages in sediments are in£uenced by various degradation agents, compared to their continental equivalents, due to transportation over greater distances. However, they can be used as a powerful tool for determining the regional climate and paleoclimatic evolution, and also as a time control beyond the limit of radiometric dating (RossignolStrick and Paterne, 1999). Pollen concentrations are generally high in sapropels in the eastern Mediterranean. The characteristic pollen assemblages often found in sapropels are dominated by the moisture-demanding trees and herbs, such as Quercus, Cedrus, Fagus, Abies, (Cheddadi et al., 1991; Cheddadi and Rossignol-Strick, 1995;

0025-3227 / 02 / $ ^ see front matter 0 2002 Elsevier Science B.V. All rights reserved. PII: S 0 0 2 5 - 3 2 2 7 ( 0 2 ) 0 0 3 4 1 - 9

MARGO 3156 3-10-02

36

H. Caner, O. Algan / Marine Geology 190 (2002) 35^46

Aksu et al., 1995; Rossignol-Strick and Paterne, 1999; Grega et al., 2000), and are associated with interglacial conditions. The abundance of resistant cold-arid herbs, such as Artemisia and Chenopodiaceae, re£ect dry and cold climate of glacial conditions. The sediment deposits in the Marmara Sea, which connects to the Black and Mediterranean seas by the Turkish Strait System (TSS) (Bosphorus and Dardanelles) are excellent tools for determining and understanding the water exchanges between two marine realms during Late Quaternary in relation to paleoceanographical evolution in a broader view. Previous investigations indicate the occurrence of sapropelic layers (CFagflatay et al., 1999, 2000) in the Marmara Sea; one is partly contemporaneous with the S1 sapropel unit of the eastern Mediterranean. The formation of these sapropelic layers is believed to be closely related with the large freshwater in£ux from the Black Sea and establishment of the two-way £ow regime (CFagflatay et al., 2000). The Late Quaternary pollen records and their paleoclimatic signi¢cance have been studied in continental sedimentary sequences around the Marmara Sea (Bottema and van Zeist, 1990; Bottema et al., 1994; Beug, 1967). Previous palynological investigations were locally carried out in sediments from the Golden Horn estuary (Ediger, 1990; Kutluk, 1994; Caner, 1994) and Izmit Bay (east of the Marmara Sea) (Akgu«n, 1995), however no previous data is available for the deep marine sedimentary deposits of the Marmara Sea. This paper presents the pollen records of the Marmara Sea sediment deposited during Late Glacial to Holocene, in relation to paleoclimatic conditions for sapropelic layers.

2. Sedimentology and paleoceanography of the Marmara Sea during Late Glacial to Holocene The intracontinental Marmara Sea (Fig. 1) is a unique waterway between Black and Mediterranean seas via TSS, and consists of various types of sedimentary depositional environments, such as shelves, slopes, deep basins and ridges. The shelf area is subject to accumulation of mainly coarse-

grained siliciclastic terrigenous sediments with biogenic sediment contribution, particularly in the southern part where it widens and receives large riverine input (Ergin et al., 1991, 1997; CFagflatay et al., 1996). Fine-grained sediment deposition prevails in the slopes and deep basinal areas, in various transportation modes, including mass movement and pelagic types. The sediment deposition and paleoceanography of the Marmara Sea during Late Quaternary has been mainly controlled by the changing sea level of the two marine realms (Black and Mediterranean seas). The permanent two-layered water strati¢cation £owing in opposite directions characterizes the º nlu«ata et al., 1990). present-day oceanography (U Sill depths of the TSS (V35 and V60 m for the Bosphorus and V70 m for the Dardanelles) constitute a substantial factor that control the water exchanges through Marmara Sea in relation to the global sea level changes and paleoceanographical implications. The e¡ect of changing sea levels of the Black and Mediterranean seas into the Marmara Sea has been limited with these sill depths (Smith et al., 1995; Ryan et al., 1997; Aksu et al., 1999; Algan et al., 2001) during the Late Quaternary. The Marmara Sea was an isolated lake during the Late Glacial to 12 kyr BP with the lowering of global sea level, as indicated by the deposition of Unit 2 in the deep basins (CFagflatay et al., 2000). Unit 2 is de¢ned with alternating light green^ gray, laminated mud with FeS reduction bands (Fig. 2). This unit in core DM13 includes an ash layer which is correlated with an 18-kyr-old Y-2 ash layer (Keller et al., 1978; CFagflatay et al., 2000) of Santorini eruption at 280 cm. The shelf areas were exposed to subaerial erosion and subject to delta progradations (Ergin et al., 1997; Aksu et al., 1999). Global sea level started to rise from ca. 3120 m at 18 kyr BP (Fairbanks, 1989), but could not reach the Marmara Sea due to the sill depth of Dardanelles (370 m) (Aksu et al., 1999). The onset of Mediterranean waters occurred at about 12 kyr BP, and the shelf areas of the Marmara Sea were submerged by the Mediterranean waters, as indicated by its typical benthic foraminifera and mollusk fauna at the base of Unit 1 (CFagflatay et al., 2000). Unit 1 consists of green^

MARGO 3156 3-10-02

H. Caner, O. Algan / Marine Geology 190 (2002) 35^46

37

Fig. 1. Location map of the study area, showing core samples sites and the present-day bathymetry.

gray to dark gray homogeneous mud, including two sapropelic layers (Fig. 2). Sapropelic layers are represented by a relatively dark, olive green^ gray mud with organic carbon content s 1.5%. 14 C age determinations indicate that the deposition of the upper and lower sapropelic layers in Unit 1 occurred between 3.2^4.7 and 6.4^10.6 kyr BP, respectively (CFagflatay et al., 2000). Following the onset of Mediterranean waters, the large fresh water in£ux from the Black Sea caused the suboxic bottom conditions resulting in sapropel deposition during 10.6^6.4 kyr BP. The lower sapropelic layer roughly corresponds to the S1 sapropel in the eastern Mediterranean, whereas the upper sapropelic layer is considered to be re£ecting the establishment of present-day two-way £ow regime (CFagflatay et al., 2000).

3. Vegetational source area of pollen The Marmara Sea is surrounded by Anatolia and Thrace peninsulas that are covered by various types of vegetation. The high variation of vegetation is caused by the geomorphic properties, climate and climatic changes during the Late Quaternary in Turkey (Zohary, 1973). The mountain chains extending along the coast of Turkey, large plains and low and high plateaus in the Central and Eastern Anatolia, together establish a transitional region between temperate and tropical belts, and continental part of Asia to the Eastern Mediterranean region (Atalay, 1994). The di¡erences in elevation among the mountain chains permit the settlement of di¡erent £oras, such as the dominance of Euro-Siberian plant elements in

MARGO 3156 3-10-02

38

H. Caner, O. Algan / Marine Geology 190 (2002) 35^46

Fig. 2. Lithological description of the cores together with the total pollen sum (counts). Core DM13 is from CFagflatay et al. (2000), and combined with the pollen data of this study.

the high altitude, and Mediterranean elements at the lower altitude of the North Anatolia Mountain (Zohary, 1973; Cohen, 1970). The main climate types prevailing in Turkey are mild-humid (oceanic), cold-humid and Mediterranean (ErincF, 1949, 1984). The Marmara Sea and its surrounding is a transitional zone between the Black Sea and Mediterranean type of climate and vegetation. In the northern part of the Marmara region, mainly Quercus (Oak), Alnus, Acer, Corylus, Ulmus, Fraxinus, Tilia occur (Yalt|r|k, 1966; Zohary, 1973; Davis, 1965^1985) as they characterize the upland broad-leaved deciduous forest of lowland Central Europe in the upper altitudinal belt, up to 2000 m (Cheddadi et al., 1991). The southern Marmara region contains two types of forests: broad-leaved and dry forest (Mediterra-

nean shrub and Pinus brutia L. communities), such as Quercus, Olea, Alnus and Pinus. The climatic conditions at 18 kyr BP were cold and arid in the central part of Anatolia, but cold and to some extent humid in the coastal areas. The temperature was 6^8‡C lower in this period than at present (van Zeist and Bottema, 1991). Anatolia was covered extensively by steppe and desert^ steppe vegetation, particularly dominated by Artemisia. The shift from the glacial to interglacial conditions is marked by the extension and migration of forests throughout the Anatolia region. Coniferous forests have migrated to the upper part of the mountains, replacing the deciduous forests (Bottema et al., 1994). The main characteristic of the vegetation in Anatolia has not substantially changed since the last 8 kyr BP, and

MARGO 3156 3-10-02

H. Caner, O. Algan / Marine Geology 190 (2002) 35^46

the boundaries of the present-day forests and steppes have formed at 4^4.5 kyr BP (Zohary, 1973; van Zeist and Bottema, 1991; Bottema et al., 1994).

4. Materials and methods Pollen analysis was carried out on two cores collected from the Marmara Sea (Fig. 1). The gravity cores DM13 and KL97 were collected during the cruises of R/V Sismik 1 and R/V Meteor, respectively. The sediment samples for pollen analysis (V1 cm3 fresh material) were collected at 10-cm intervals in both cores and treated using standard palynological techniques (Erdtman, 1954; Moore et al., 1991). This method includes HCl, HF, and KOH digestion, before staining with safranine and mounting with glycerine jelly. The pollen percentages are based on the pollen sum of arboreal (AP) and non-arboreal pollen (NAP), excluding spores. The number of total pollen grains counted per each sample was about 1500^1600 and normalized with the amount of slide scanned.

5. Results and discussion 5.1. Sediments The core DM13 was collected from the western ridge of the Marmara Sea at a water depth of 710 m, with a recovery of 3 m. The lithological and chronological description of this core is summarized under 2. Sedimentology and paleoceanography of the Marmara Sea during Late Glacial to Holocene and illustrated in Fig. 2. The highest amount of total pollen coincides with the lower sapropelic layer (Fig. 2). KL97 was collected from the Eastern Basin of the Marmara Sea at a water depth of 1094 m. It has a length of 540 cm and consists of mainly green^gray mud with local laminations and ¢negrained turbiditic sand bands. The organic carbon content of the sediments in this core £uctuates between 0.3 and 3.1%, and is relatively constant throughout the core, with the exception of two

39

peaks of 1.5^3.1%, between 110^130 cm and 230^260 cm, coinciding with sediment color changes from gray to green^gray and dark gray (Fig. 2). This core was recovered from a location which is very close to the slope^basin boundary according to recent detailed bathymetrical information (Gaziog›lu et al., 2002), and is likely susceptible to high sedimentation rate. Previous sedimentological studies (Stanley and Blanpied, 1980; Evans et al., 1989; CFagflatay et al., 2000) in the Marmara Sea revealed highly variable sedimentation rates in the deep basins, during the Last Glacial to Holocene. Sedimentation rates based on 14 C ages for core DM13 were estimated to increase from 0.1 to 0.25 m/kyr between Unit 1 and Unit 2 (Fig. 2; CFagflatay et al., 2000). The 18kyr-old ash layer presented in core DM13 was also observed in a nearby core (KL40, 703 m water depth) at V600 cm downcore, which suggests highly variable sedimentation rates in this core (Wulf et al., 2002). Due to these high variations of sedimentation rates even at a very short distance in the Marmara Sea, the time scale was not extrapolated with the available dating data, to avoid misinterpretation. Therefore a stratigraphical correlation between the cores DM13 and KL97 was attempted using the sapropelic layer intervals and pollen zones. The pollen assemblage zones (see below) are also well correlated within these two cores. On the whole, the sedimentation rate for KL97 seems to be twice the rate for core DM13. The lack of radiometric dating and possible instantaneous (and/or disturbed) deposition pattern in core KL97 might be considered as a substantial restriction for the stratigraphic correlation of the two sites. The palynological stratigraphy is consistent between the two cores, and the pollen records in KL97 display a signi¢cant distribution pattern that could not have been formed under mixed or disturbed conditions. 5.2. Pollen assemblage zones The pollen records of the sediment cores include several types of vegetation, from steppe to semi-desert taxa, to those of upland broad-leaved deciduous and coniferous forests. Quercus is the most common tree pollen (10^50% of the £ora) in

MARGO 3156 3-10-02

40

H. Caner, O. Algan / Marine Geology 190 (2002) 35^46

Fig. 3. Distribution of total pollen and relative abundance of pollen species in core DM13. Total pollen denotes the counts (sum) for each sampling interval. AP/NAP graph is based on percentages in pollen sum.

both cores, re£ecting the eastern Mediterranean warm-dry summers and mild-wet winters (Davis, 1965^1985). Pinus is the second most abundant pollen taxa in core KL97, but is unexpectedly very scarce in DM13. Steppe-type pollen, such as Artemisia, mountain evergreen taxa, and Picea

are only found in abundance in core KL97. The downcore variations of total AP and NAP percentages, together with cold-arid and warm-wet climate-demanding individual pollen taxa are shown in Figs. 3 and 4. Four di¡erent zones are distinguished along the cores based mainly on the

Fig. 4. Distribution of total pollen and relative abundance of pollen species in core KL97. Total pollen denotes the counts (sum) for each sampling interval. AP/NAP graph is based on percentages in pollen sum.

MARGO 3156 3-10-02

H. Caner, O. Algan / Marine Geology 190 (2002) 35^46

variations of AP and NAP percentages, and also on the changing abundance of individual tree and herb taxa. 5.3. DM13 The highest total pollen count between 80 and 100 cm, which reaches about 900, is coeval with the lower sapropelic layer in this core (Fig. 3). The total AP percentages are relatively higher than the NAP throughout the core. Zone D at the base of the core (from 200 to 290 cm) is characterized by a high value of AP, ranging between 80 and 90%, and consequently low value of NAP. Quercus is the dominant AP, followed by irregularly increasing Pinus, Juniperus, Alnus, Fraxinus, Taxus and Corylus. Chenopodiaceae is the most abundant NAP, with a maximum of 15% in this zone. Between the 120^190-cm intervals, in Zone C, the amount of total AP increases against NAP (up to 91%). The percentage of Quercus does not show variation, while Pinus, Castanea and Alnus are decreasing, and Carpinus, Taxus and Corylus are increasing. Chenopodiaceae is the dominant NAP, but displays a decreasing trend at the base of the Zone C. The increasing pollen sum de¢nes the Zone B. This zone is located between 60 and 120 cm of the core and corresponds to the lower sapropelic layer. Quercus, Pinus, Juniperus, Castanea, Alnus, Olea, Ulmus and Ostrya increase to their maximum values, whereas Colchicum, Rumex and Chenopodiaceae start to decrease. Zone A is de¢ned by almost constant distribution of AP and NAP percentages with the exception of a slight increase of AP value at the top of the core. Relatively noticeable variations are observed in increasing Plantago, and decreasing Polygonum, Rumex, Pinus, Alnus, Olea, Ulmus, Fraxinus, Carpinus and Corylus. Olea displays an increment, but decreases towards the surface. The percentages of Chenopodiaceae and Rumex are the lowest among the NAP. 5.4. KL97 The total pollen sum (or pollen count) displays

41

the highest value at around 260 cm, corresponding to the sapropelic layer (Fig. 4). Zone distinctions are more clearer and pollen percentages are more consistent in this core, compared to those of DM13. Zone D, located between 400 and 540 cm, is characterized by increasing abundance of AP values and consequently decreasing NAP values. AP values £uctuate between 57 and 72% with slightly increasing Pinus and Juniperus. Quercus displays a decreasing trend in this zone, together with Olea and Corylus. Within the NAP assemblage, Artemisia, Chenopodiaceae and Plantago are the most abundant pollen taxa. Chenopodiaceae and Plantago reach their highest values. Zone C of core KL97 (280^400 cm) de¢nes the highest percentage of AP and lowest percentage of NAP throughout the core. AP values increase to their maximum value of 79%, at about 340 cm. The AP assemblage is mainly characterized by the dominance of Pinus accompanied by Picea, Olea, Corylus and Ulmus. Although the total NAP values decrease in this zone, some of the species, such as Artemisia, Colchicum and Lilium, reach their maximum values. Chenopodiaceae and Plantago display a decreasing trend. In contrast to Zone C, AP percentages decrease, while NAP percentages increase in Zone B, in between the 140^280-cm interval of the core KL97. Several AP species decrease (Pinus, Picea and Olea) or disappear (Ulmus) in that zone, however, Quercus, Castanea and Corylus signi¢cantly increase to their maximum value, together with the ¢rst appearance of Alnus and Acer. Zone A de¢nes slightly increasing AP and decreasing NAP values without any major £uctuations of pollen percentages, except Tilia, Quercus and Alnus. These pollen grains increase in upward direction. Acer shows a minimum between 40 and 80 cm. 5.5. Implications for paleoclimatology Both cores from the Marmara Sea contain four palynological zones with their characteristic pollen distribution pattern. However, pollen assemblages display slightly di¡erent composition in the east and west part of the Marmara Sea. KL97

MARGO 3156 3-10-02

42

H. Caner, O. Algan / Marine Geology 190 (2002) 35^46

from the eastern deep basin includes sagebrush Artemisia and the moisture-demanding mountain tree, Picea, which are absent in DM13. On the other hand, DM13 has a greater diversity in typical Mediterranean temperate character of AP (Fraxinus, Taxus, Similax, Juglans and Ostrya), compared to KL97. The highest total pollen sum along the cores corresponds to the sapropelic layers due to the favorable preservation condition during bottom water stagnation. Variation of pollen production or transportation cannot account for the high pollen concentration in a marine depositional system, since these factors are not subject to great variations (Rossignol-Strick, 1985; Cheddadi et al., 1991; Cheddadi and RossignolStrick, 1995). On the basis of 14 C ages in DM13 and the location of sapropelic layers in both of the cores, Zone B can be attributed to the deglaciation (or Interglacial) period. The lower two zones must be re£ecting the preceding Late Glacial period. This rough division and dating is in agreement with the characteristic pollen assemblage distribution and discussed below. Pollen distribution in the deep Marmara Sea sediments is mainly dominated by diversi¢ed AP, suggesting that the source area was characterized by altitudinal-controlled vegetational belts. The upland broad-leaved conifer and deciduous tree types, such as Juniperus, Pinus, Castanea, Corylus and Ulmus, together with the dominance of Quercus in Zone D of the two cores re£ects the Mediterranean climate, which is mild and humid in winter, and warm and dry in summer. Their low abundance compared to those of the upper zones might be related to the reduced humidity. The dominance of Chenopodiaceae (core DM13) together with Artemisia (core KL97) among the NAP in this zone reveals the open steppe vegetation, which characterizes very arid and continental conditions. During the Late Glacial, Artemisia and Chenopodiaceae were largely represented in the terrestrial and lacustrine depositional environments of southwest and northern Turkey (Bottema et al., 1994), northwest Greece and Central Europe (van Zeist et al., 1975), as a result of prevailing steppe conditions. Zone C of core KL97 is marked by the NAP assemblage with the highest values of Artemisia (up to 50%), Col-

chicum and Lilium. Artemisia is an indicator for fully glacial periods with cold and dry conditions (Cheddadi et al., 1991; Rossignol-Strick and Paterne, 1999). Likewise, Quercus percentage, the indicator of opposite conditions, is the lowest (7%), and the other tree pollen grains decrease or almost become absent, with the exceptions of Olea, Pinus, Ulmus and emerging of Picea. The low abundance of Juniperus, Castanea and Alnus may imply their reducing population in the source area because of migrating to cold and humid coastal areas. Changing conditions are indicated with strong £uctuating and relatively decreasing values of Chenopodiaceae and other herbs in DM13, for the base of the same zone. Picea is a conifer tree present in Central Europe and North Anatolia Mountains of Turkey and Caucasian Mountains at present. It started to occupy in the North Anatolia Mountains after 14 kyr BP (Zohary, 1973; Bottema et al., 1994). The NAP assemblage of Chenopodiaceae, Artemisia and Plantago suggests a cold and dry source area that might be high plateaus of Turkey and southeast Russia from the steppe vegetation (Wright et al., 1967). However, from zones D to C of core KL97, Artemisia starts to increase, while Chenopodiaceae are still abundant but display a decreasing trend. Chenopodiaceae are a more arid herb than Artemisia. The presence of Olea also indicates the in£uence of Mediterranean-type climate. The climate must have been cold and more arid in Zone D and gradually has changed to relatively humid conditions at the end of Zone C. In contrary to Zone C, Zone B is presented with the increasing moisture-demanding AP assemblages, including Quercus, Pinus, Juniperus, Castanea, Alnus, Olea, Ulmus and Ostrya in core DM13, and appearances of Acer and Alnus in core KL97. The abundance of Quercus in both cores indicates the warm and wet climate of deglaciation. The indicator of opposite conditions, Artemisia, is absent throughout the core DM13 and decreases in this zone of core KL97. Such distribution of pollen grains in Zone B suggests that the deciduous forests have greatly expanded with the regression of steppe vegetation, as previously indicated by Bottema et al. (1994), north-

MARGO 3156 3-10-02

H. Caner, O. Algan / Marine Geology 190 (2002) 35^46

east of Anatolia. The deposition of lower sapropelic layer corresponds to this zone and is also evident from the increasing peak of the total pollen sum along both of the cores. These temperate moisture-demanding tree pollen grains are abundant in the Eastern Mediterranean recent sapropel S1 and synchronous with the climatic optimum (ca. 10^5.7 kyr BP; Cheddadi et al., 1991). Alnus and Acer in core KL97 appear in this zone. Acer is absent in core DM13, however, Alnus has its maximum value in Zone B of the two cores. The abundance of Alnus in the Black Sea sediments has been related to warming conditions during the sea level rise (Traverse, 1974). Acer is also replaced by Betula forest during 10^7 kyr BP in the northwest of Turkey (Bottema et al., 1994). Hence, the sudden appearance of Alnus and Acer, together with increasing Tilia that characterize the Balkan and Caucasian mountains and European lowlands (Rossignol-Strick, 1973), might be attributed to a Black Sea source area. Similarly, the deposition of lower sapropelic layer in the same zone of DM13 has been initiated during a large out£ow from the Black Sea (CFagflatay et al., 2000; Aksu et al., 1999). Although chronological dating is absent in core KL97, the depositions of sapropelic layers probably have a common origin in both cores. The warmer climate conditions indicated by the pollen assemblage in this zone corresponds to 12 kyr BP with a close estimate from the age data of DM13. This is almost the end of glaciation period and also corresponds to the Mediterranean waters entering through the Dardanelles sill. Climate warming occurred after the Last Glacial Maximum at 18 kyr BP, when the Marmara Sea was disconnected from the Mediterranean. The maximum occurrence of Artemisia and decreasing trend of Chenopodiaceae together with Olea, Pinus, Ulmus and Picea in Zone C might be re£ecting the in£uence of warming climate whereas Zone D may closely be associated with the glacial conditions. The distribution patterns of the most dominant AP assemblages, such as Quercus, Juniperus and Castanea, do not show signi¢cant £uctuation, whereas the lesser abundant species decrease towards the surface in Zone A of core DM13. The majority of the pollen grains in the same zone of

43

core KL97 displays almost unvarying distribution with the exception of few increasing species. This unvarying behavior of AP and NAP assemblages probably indicates the last 4^5 kyr to present conditions. The present-day vegetation and the extent of forest areas have not changed for the last 4 kyr on the basis of pollen studies in the Marmara region and Anatolia (Beug, 1967; van Zeist and Bottema, 1991; Bottema and van Zeist, 1990; Bottema et al., 1994; Kutluk, 1994). Although Pinus is widely known as the most common pollen in marine sediments (Cheddadi and Rossignol-Strick, 1995; Rossignol-Strick and Paterne, 1999), it is remarkably low in the deep basin sediments from the Marmara Sea. Pinus percentages are usually high in the continental sedimentary sequences around the Marmara region (Caner, 1994; Bottema et al., 1994), Golden Horn estuarine sediment (Kutluk, 1994; Ediger, 1990) and Izmit Bay (Akgu«n, 1995). Marine sediments from the Aegean Sea and Black Sea contain up to 50% Pinus (Aksu et al., 1995; Roman, 1974). Pinus brutia L. was recognized to be the most common pollen type since the last 8 kyr in the lacustrine sediments from the southern Marmara region (Bottema et al., 1994). The morphology of Pinus grains is distinctive with two separate sacci and therefore can be transported to longer distances compared to other bi-saccatetype pollen grains (Aytug›, 1969). Salinity (density) of water mass is also important in transportation or deposition of saccate pollen grains as they can remain suspended for a longer period of time in marine water, compared to fresh water (Traverse and Gingsburg, 1966). The low occurrence of Pinus in the Marmara Sea sediments might be attributed to the density di¡erences in two-layer strati¢cation of the water column. The interface layer between less saline Black Sea origin waters and more saline Mediterranean origin waters, permanently exist throughout the year with a thickness of 10^20 m (BesFiktepe et al., 1994), and acts as a trap for particulate material. From this point of view, Pinus pollen grains originating from both atmospheric and riverine inputs might be trapped in the interface layer and mainly transported from the Marmara Sea to Mediterranean due to their morphology and buoyancy.

MARGO 3156 3-10-02

44

H. Caner, O. Algan / Marine Geology 190 (2002) 35^46

The absence of Picea and Artemisia in core DM13 remains unexplained in this study. The resultant factors can be variable, such as variation of dynamic oceanographical conditions of the Marmara Sea during the Late Glacial^Holocene and also relatively increasing in£uence of local source area. The presence of Picea and Artemisia in KL97 and their absence in core DM13 might be attributed to the proximity to their source area (North Anatolia^Caucasia^Central Europe^Siberia), although it is unlikely to expect such variation in a short distance (the distance between the two cores is about 80 km). Furthermore, none of these two pollen species have been found in Izmit Bay (Akgu«n, 1995) and Golden Horn (Ediger, 1990), close to the locality of KL97. The dynamic oceanography of the Marmara Sea might be a factor for this unusual pollen dispersal. The upper layer entering from the Bosphorus £ows to the south initially, then inclines to the west forming an ‘S’-shaped jet £ow. During autumn and winter, the jet becomes weaker due to low discharges from the Black Sea, and have a tendency to £ow over the northern shelf near to exit from the Bosphorus (BesFiktepe et al., 1994). Close to the exit from the Bosphorus, the decrease in velocity of the upper layer originating in the Black Sea can explain the restricted presence of Artemisia and Picea. However, this explanation is based solely on the assumption of their transportation mainly by water from the Black Sea and ignores the atmospheric input. Therefore, further palynological analysis with age determination in di¡erent parts of the Marmara Sea should explain this aspect with reliable evidences and also might help to detect the short-term climatic variations (like Younger Dryas).

and Chenopodiaceae in core DM13. Increasing amounts of some deciduous and coniferous tree pollen grains and a decreasing trend of Chenopodiaceae in Zone C re£ect the beginning of warm and wet conditions. Zone B corresponds to prevailing fully interglacial conditions with warmer and more humid climate indicated by the diversi¢ed Mediterranean temperate pollen taxa. The total pollen sum in this zone reach to their maximum especially in the sapropelic layers. The appearance and increasing percentage of Alnus and Acer together with Tilia in this zone of core KL97 might be attributed to input from the Black Sea region, although they are absent in core DM13. Pollen assemblages of the Zone A do not show a signi¢cant variation.

Acknowledgements We thank the o⁄cers, crew and scienti¢c personnel of the R/V Sismik 1 of the General Directorate of the Mineral Research and Exploration (MTA) and R/V Meteor of the Federal Republic of Germany. We gratefully acknowledge Prof. P. Halbach (Berlin Freie University) for providing the core KL97 during the cruise M44/1, and Prof. B. Aytug› (Istanbul University) for his enlightening comments and discussions on the pollen records. Special thanks are due to A. AytacF (Msc. Student) and E. SarIW (Ph.D Student) for their help during preparation of pollen slides and organic carbon analysis. We are grateful to Profs. A. Rochon, M.B. Cita and S. Leroy for their critical and constructive reviews that signi¢cantly improved the manuscript. References

6. Conclusions Pollen records of the deep basin sediments from the Marmara Sea reveal the existence of four different palynological zones related to the Last Glacial/Interglacial paleoclimatic changes. Zone D indicates prevailing cold and arid conditions in the source area, as re£ected by the increasing steppetype herbs, such as Artemisia in core KL97

Akgu«n, F., 1995. IWzmit Ko«rfezi Kuaterner isti¢nin palinolojik incelemesi. In: MericF, E. (Ed.), Izmit Korfezi Kuaterner Isti¢. Kocaeli Valiligfli CFevre Koruma Vakf|, Istanbul, pp. 179^201 (in Turkish with English abstract). Aksu, A.E., YasFar, D., Mudie, P.J., Gillespie, H., 1995. Late glacial-Holocene palaeoclimatic and palaeoceanographic evolution of the Aegean Sea: micropaleontological and stable isotopic evidence. Mar. Micropaleontol. 25, 1^28. Aksu, A.E., Hiscott, R.N., YasFar, D., 1999. Oscillating Quaternary water levels of the Marmara Sea and vigorous out-

MARGO 3156 3-10-02

H. Caner, O. Algan / Marine Geology 190 (2002) 35^46 £ow into the Aegean Sea from the Marmara Sea^Black Sea drainage corridor. Mar. Geol. 153, 275^302. Algan, O., CFagflatay, N., Tchepalyga, A., Ongan, D., Eastoe, C., Go«kasFan, E., 2001. Stratigraphy of sediment in¢ll in Bosphorus Strait: water exchange between the Black and Mediterranean seas during the last glacial Holocene. GeoMar. Lett. 20, 209^218. Atalay, I., 1994. Vegetation Geography of Turkey. Ege Universitesi Basimevi, Izmir. Aytugfl, B., 1969. Atmosfer pollen analizleri ve bu analizlerin º niversitesi, Orman Faku«ltesi Dergisi, faydalari. IWstanbul U Ser. A, XIX, pp. 71^94 (in Turkish with French abstract). Beug, H.J., 1967. Contributions to the Postglacial vegetational history of northern Turkey. Quat. Palaeoecol. 7, 349^ 356. º zsoy, E., Latif, M.A., Ogfluz, T., BesFiktepe, SF.T., Sur, H.IW., O º nlu«ata, U º ., 1994. The circulation and hydrography of the U Marmara Sea. Prog. Oceanogr. 34, 285^334. Bottema, S., van Zeist, W., 1990. Middle East Early Holocene Vegetation (ca. 8000 BP). Ludwig Riechert Verlag, Wiesbaden. Bottema, S., Woldring, H., Aytugfl, B., 1994. Late Quaternary vegetation history of Northern Turkey. Palaeohistoria 35^ 36, 13^72. CFagflatay, M.N., Algan, O., Balk|s, N., Balk|s, M., 1996. Distribution of carbonate and organic carbon contents in Late Quaternary sediments of the southern Marmara shelf. Turk. J. Mar. Sci. 2, 67^83. CFagflatay, M.N., Algan, O., Sak|ncF, M., Eastoe, C.J., Egesel, L., Balk|s, N., Ongan, D., Caner, H., 1999. A mid-late Holocene sapropelic sediment unit from the southern Marmara Sea shelf and its palaeoceanographic signi¢cance. Quat. Sci. Rev. 18, 531^540. CFagflatay, M.N., Go«ru«r, N., Algan, O., Eastoe, C., Tchapalyga, A., Ongan, D., Kuhn, T., KusFcu, IW., 2000. Late glacial-Holocene palaeoceanography of the Sea of Marmara: timing of the last connections with the Mediterranean and the Black seas. Mar. Geol. 167, 191^206. Caner, H., 1994. IWstanbul’da KentlesFmenin Dogflal Orman Alanlar|na Etkilerinin Dendrokronoloji ve Palinoloji Yo«ntemleri ile Belirlenmesi (in Turkish). Unpubl. PhD Thesis, Institute of Marine Sciences and Geography, Istanbul University, Istanbul. Cheddadi, R., Rossignol-Strick, M., 1995. Improved preservation of organic matter and pollen in Eastern Mediterranean sapropels. Paleoceanography 10, 301^309. Cheddadi, R., Rossignol-Strick, M., Fontugne, M., 1991. Eastern Mediterranean paleoclimates from 26 to 5 ka B.P. documented by pollen and isotopic analysis of a core in the anoxic Bannock Basin. Mar. Geol. 100, 53^66. Cohen, H.R., 1970. The paleoecology of South Central Anatolia at the end of the Pleistocene and beginning of the Holocene. Anatol. Stud. 20, 119^138. Davis, P.H., 1965^1985. Flora of Turkey and East Aegaean Islands, Vols. 1^10. Edinburgh University Press, Edinburgh. Ediger, V.S., 1990. Palinoloji. In: MericF, E. (Ed.), Istanbul Bogflazi Gu«neyi ve HalicF’in GecF Kuvaterner (Holosen) dip

45

º niversitesi Vak¢, Istanbul, Tortullari. IWstanbul Teknik U pp. 59^71 (in Turkish with English abstract). Erdtman, G., 1954. An Introduction to Pollen Analysis. Chronica Botanica. Ergin, M., Bodur, M.N., Ediger, V., 1991. Distribution of sur¢cial shelf sediments in the northeastern and southwestern parts of the Sea of Marmara: Strait and canyon regimes of the Dardanelles and Bosporus. Mar. Geol. 96, 313^340. º ., Karadenizli, L., Ergin, M., Kazanc|, N., Varol, B., IWleri, O 1997. Sea-level changes and related depositional environment on the southern Marmara shelf. Mar. Geol. 140, 391^403. ErincF, S., 1949. The climates of Turkey according to Thornthwaite’s classi¢cations. Ann. Assoc. Am. Geogr. 39, 26^46. ErincF, S., 1984. Klimatoloji ve metodlari (in Turkish). IWstanbul º niversitesi Yay|n no. 3278, Istanbul. U Evans, G., Erten, H., Alavi, S.N., Von-Gunten, H.R., Ergin, M., 1989. Sur¢cial deep-water sediments of the Eastern Marmara Basin. Geo-Mar. Lett. 9, 3^27. Fairbanks, R.G., 1989. A 17,000-year glacio-eustatic sea level record: in£uence of glacial melting rates on the Younger Dryas event and deep-ocean circulation. Nature 342, 637^ 642. Gaziogfllu, C., Algan, O., Go«kasFan, E., Yu«cel, Z.Y., Tok, B., Dogflan, E., 2002. Morphologic features of the Marmara Sea from the multi-beam data. Mar. Geol. 190, S00253227(02)00356-0. Grega, M., Tsalia-Monopolis, St., Ioakim, C., Papatheodorou, G., Ferentinos, G., 2000. Evaluation of palaeoenvironmental changes during the last 18,000 years in the Myrtoon basin, SW Aegean Sea. Palaeo 156, 1^17. Keller, J., Ryan, W.B.F., Ninkovich, D., Altherr, R., 1978. Explosive volcanic activity in the Mediterranean over the past 200,000 yr as recorded in the deep sea sediment. Geol. Soc. Am. Bull. 89, 591^604. Kutluk, H., 1994. HalicF Holosen Polenleri (in Turkish). Unpubl. PhD Thesis, Institute of Marine Sciences and Geography, Istanbul University, Istanbul. Moore, P.D., Webb, J.A., Collinson, M.E., 1991. Pollen Analysis. Blackwell Scienti¢c Publications, Oxford, 216 pp. Roman, S., 1974. Palynoplanktologic analysis of some Black Sea cores. In: Degens, E.T., Ross, D.A. (Eds.), The Black Sea ^ Geology, Chemistry, and Biology. Am. Assoc. Pet. Geol. Mem. 20, pp. 396^411. Rossignol-Strick, M., 1973. Pollen analysis of some sapropel layers from the deep sea £oor of the eastern Mediterranean. Init. Rep. DSDP 13, 971^991. Rossignol-Strick, M., 1985. Mediterranean Quaternary sapropels, and immediate response of the African monsoon to variation of insolation. Palaeogeogr. Palaeoclimatol. Palaeoecol. 49, 237^263. Rossignol-Strick, M., Paterne, M., 1999. A synthetic pollen record of the Eastern Mediterranean sapropels of the last 1 Ma: implications for the time-scale and formation of sapropels. Mar. Geol. 153, 221^237. Ryan, W.B.F., Pitmann, W.C., III, Major, C.O., Shimkus, K.,

MARGO 3156 3-10-02

46

H. Caner, O. Algan / Marine Geology 190 (2002) 35^46

Moskalenko, V., Jones, G.A., Dimitrov, P., Go«ru«r, N., Sak|ncF, M., Yu«ce, H., 1997. An abrupt drowning of the Black Sea shelf. Mar. Geol. 138, 119^126. Smith, A.D., Taymaz, T., Oktay, F., Yu«ce, H., Alpar, B., BasFaran, H., Jackson, J.E., Kara, S., SFimsFek, M., 1995. High-resolution seismic pro¢ling in the Sea of Marmara (northwest Turkey): Late Quaternary sedimentation and sea-level changes. Geol. Soc. Am. Bull. 107, 923^936. Stanley, D.J., Blanpied, C., 1980. Late Quaternary water exchange between the eastern Mediterranean and the Black Sea. Nature 285, 537^541. Traverse, A., 1974. Palynological investigation of two Black Sea cores. In: Degens, E.T., Ross, D.A. (Eds.), The Black Sea ^ Geology, Chemistry, and Biology. Am. Assoc. Pet. Geol. Mem. 20, pp. 381^389. Traverse, A., Gingsburg, R.N., 1966. Palynology of the surface sediments of great Bahama Bank, as related to water movement and sedimentation. Mar. Geol. 4, 417^459. º zsoy, E., 1990. On the º nlu«ata, U º ., Ogfluz, T., Latif, M.A., O U physical oceanography of the Turkish Straits. In: Pratt, L.J. (Ed.), The Physical Oceanography of Sea Straits. Kluwer Academic Publishers, Dordrecht, pp. 25^60.

Yalt|r|k, F., 1966. A study on the £oristic analysis of vegetation of the Belgrad Forest and composition of the main stand types. Tarim Bakanligi Orman Genel Mu«du«rlu«gflu« Yayin 436^6, Istanbul. van Zeist, W., Bottema, S., 1991. Late Quaternary vegetation of the Near East. In: Beihefte zum Tu«binger Atlas des Vorderen Orients, Reihe A: Naturwissenschaften 18. Ludwig Reichter, Wiesbaden. van Zeist, W., Woldring, H., Stapert, D., 1975. Late Quaternary vegetation and climate of Southwestern Turkey. Palaeohistoria 17, 55^142. Wright, H.E., Jr., McAndrews, J.H., van Zeist, W., 1967. Modern pollen rain in western Iran and its relation to plant geography and Quaternary vegetational history. J. Ecol. 55, 415^443. Wulf, S., Kraml, M., Kuhn, M., Schwarz, T., Inthorn, N., Halbach, P., Keller, J., KusFcu, IW., accepted. Marine tephra from the Cape Riva eruption (22 ka) of Santorini in the Sea of Marmara (Turkey). Mar. Geol., accepted. Zohary, M., 1973. Geobotanical Foundations of the Middle East, Vol. I, II. Gustav Fischer, Stuttgart.

MARGO 3156 3-10-02