Climate Changes During The Holocene In The Levant

The initial procedure adopted by the author to establish the sequence of climate changes during the Holocene was based on a chrono-stratigraphical cross section derived mainly from the sequence of ratios of 18O/16O, with the ratios of 13C/12C as auxiliary data (Fig. 1.2). These isotopic data came from a core from the bottom of Lake Van in Turkey (Lemcke and Sturm, 1997), from a core at the bottom of the Sea of Galilee (Stiller et al., 1983— 84), from speleothemes of caves in upper Galilee (Issar (1990) based on M. A. Geyh et al., unpublished data) and from the Soreq Cave in the Judean hills in the central part of Israel (Bar-Matthews et al., 1998a,b) and from cores at the bottom of the eastern most part of the Mediterranean Sea (Luz, 1979; Schilman et al., 2002). Needless to say, each time series has its advantages as well as constraints, especially when it comes to the dating of the various layers. Consequently, the time boundaries suggested in this cross section (Fig. 1.2) should be taken as a synthesis and a marker zone, which may fluctuate on the time dimension either because of the natural environment or because of the different methods of sampling and dating.

The reason for choosing sequences of ratios of 518O/16O (the relative proportion of 18O to 16O in the sampled water compared with the isotopic composition of standard mean ocean water (SMOW)) as the most significant time series was because these ratios are strongly influenced by the ambient temperatures and climate regimes in general (Ferronsky and Polyakov, 1982; Fritz and Fontes, 1980; Gat, 1981) but are not influenced by anthropogenic activities. It was also assumed that, in the Middle East, the influence of climate changes on the 518O/16O ratio could have been rather pronounced, based on the observation that the isotopic composition of contemporary rainwater is influencedby the trajectories of the rainstorms (Leguy et al., 1983). There is no reason to suggest that such changes in the global climate regime would not have equally influenced these trajectories, and thus the 518O/16O, in the past. Therefore, interpretation of the stable isotope data as climate and humidity indicators follows the basic assumption that the 518O values of precipitation are interrelated with temperature (Geyh and Franke, 1970) and with other meteorological factors (such as changes in the storm trajectories, in the seasonal distribution of precipitation and humidity (Gat, 1981; Leguy etal., 1983) and higher or lower rates of evaporation). This assumption was indeed justified by the interrelations that could be shown between the isotope time series and other proxy-data time series, as will be shown below.

As already mentioned, when correlation lines are drawn, small discrepancies caused by the dating and time scales used in the different data sources must be taken into consideration. These apply to the different amplitudes of the 518O records of the lake and sea sediments and of the speleothemes. For example, the water balance of the Sea of Galilee is also determined by an inflow of groundwater from the flanks of the rift valley. Spring water collected along the shore yielded 14C dates of more than 10 ka BP. This would "dampen" the corresponding isotope variations. A certain retardation factor should be taken into consideration for the isotopic composition of the sediments of Lake Van, where part of its inflow comes from springs. In contrast, changes in 518O values of speleothemes reflect the fluctuations of isotope composition of the meteoric water over decades. The samples of 1 mm thickness analyzed represent age ranges of about 10 years.

In addition to the problems involved in the 14C dates in relation to the isochrones, some other elements must be taken into consideration. First, the curves presented in the cross sections are modified by the running average method, in order to reduce the impact of noise created by short-term but intense fluctuations. Second, there are differences caused by the reservoir effect of the non-saturated and saturated zones in the subsurface of speleothemes, which is similar to the effect of groundwater storage for springs. Yet even with all these uncertainties, an apparent general

in sediments Lake Van Based on data from Lemcke& Sturm (¡997)

Levant i 5(1 -» 50 0 » SO 1.50 PDB History climate

6,h0 &"C in sediments ft"0 &"C 111 stalagmites Soreq Cave

Sea of Galilee in caves Upper Galilee Based on data from

Bar-Matthews et al. (1997)*

Stiller et al (1983-84)* Issar (1990)

i,"o%0PDB

SXKI HI'

4000 HP

7iK«) HP, tiixxi up i)'s O in foraminifers Mediterranean Sea Based on data from Schilman et al. (2002)*

6,h0 &"C in sediments ft"0 &"C 111 stalagmites Soreq Cave

Sea of Galilee in caves Upper Galilee Based on data from

Bar-Matthews et al. (1997)*

i)'s O in foraminifers Mediterranean Sea Based on data from Schilman et al. (2002)*

SXKI HI'

4000 HP

Fig. 1.2. Time series for environmental isotopes in the Middle East. 1 Adjusted to scale and streamlined (3-5) points by the running average method.

correlation of covariations can be observed. However, because of the problems outlined above, it is suggested that conclusions should also take into consideration other time series of natural proxy-data that are available for this region. These include the paleo-levels of the Mediterranean Sea, the ratios of planktonic foraminifers in the sediments of the eastern Mediterranean, and the Dead Sea lake levels (Fig. 1.3).

As presented in Fig. 1.2, the cross section starts with the 518O and 513C time series obtained from lacustrine carbonate cores drilled in Lake Van in Eastern Turkey (Lemcke and Sturm, 1997; Schoell, 1978), which is a closed lake at an altitude of 1720 m above MSL. The precipitation on the drainage basin of the lake is influenced by the Mediterranean climate system. The isotopic investigation is part of a general study that has been carried out by a multidisciplinary group (Degens etal., 1984.) The lake has a volume of 607 km3 and a maximum depth of 451 m and is in a tectonically active zone in eastern Anatolia. The lake level was at its highest at the height of the Last Ice Age, about 18 ka BP when it was 72 m above the present level. According to the pollen analysis, the vegetation was of a steppe type from 10 ka to 6.5 ka BP; from 6.5 ka to 3.4ka BP, it was forest vegetation and from 3.4ka BP to the top of the section, the vegetation is contemporary and shows the impact of agriculture. The drop in the level of the lake and the increase in the salinity of the water between 10 and 9 ka BP were interpreted as a change to a warmer and dryer climate. This can be observed in a trend towards a heavier composition of the 18O/16O ratios in the isotope curve. Around 7ka BP, there was a rise in the level of the lake, a decrease in its salinity and a marked increase in the percentage of arboreal pollen. This is interpreted as a change to a more humid climate. One can observe a simultaneous decrease in the 18O/16O ratios. At c. 3.5 ka BP there is again a sharp decrease in the 18O/16O ratio, which reaches its lowest level at 2.7 ka BP and marks another cold period. Because of the increase in agricultural activities since then, the pollen and sed-imentological records may present the impact of anthropogenic processes, and the author prefers to rely mainly on the isotope curve, which shows relatively low ratios from 1.8 ka to c. 0.8 ka BP, a heavier composition between 0.8 ka and 0.5 ka BP and an increase in the ratio at the top of the column.

Another isotopic composition time series, presented in Fig. 1.2, is that from a core taken from the Sea of Galilee (Stiller et al., 1983-84). This lake is fed by the Jordan River, and by the floods and springs from Galilee and the Golan Heights. Thermal springs also flow into the lake. The base flow of the Jordan is maintained by the outflow of springs emerging from the aquiferous Jurassic limestone rocks of Mount Hermon, in the eastern part of the Anti-Lebanon. These rocks are highly permeable and the water from the rain falling on the mountain and from the melting snow, which covers the higher stretches of the mountain each winter, quickly infiltrates the subsurface to enrich the aquifer from which these springs arise. The average annual precipitation on the mountains may reach 1200 mm. The two main springs feeding the upper Jordan are the Dan and the Banias (comes from Pan, the Greek god patron of springs). Because of high permeability and the high rate of precipitation, the water flow of these two major springs is fairly regular. The difference between summer and winter is regulated by the large underground storage of Mount Hermon. A long spell of dry years and low snowfall on the drainage basin may cause a decrease in the total quantity of water in the springs, leading to a reduction in the flow of the Jordan and a low water level in the Sea of Galilee. This is intensified by a decrease in the volume of the floods and by higher evaporation rates from the lake, causing the levels of the lake to drop. One may assume that the ratio of 18O/16O in the carbonate sediments will be higher in such years. While the precipitation on the catchment area of the springs emerging from the southern tip of Mount Hermon is high, the precipitation on eastern Galilee and the Golan Heights, which form the catchment area of the floods and springs, is less abundant, and the rates of flow are strongly influenced by the average annual rainfall.

The reinterpretation, carried out by the present author, of the 518O/16O sequence from this core was correlated with the 513C/12C, data, assuming that depleted ratios signify more humid conditions, and thus abundant C3 types of vegetation, while a heavier composition indicates a drier climate and abundance of C4 type of vegetation.

Only four 14C dates (at 5240 ± 520, 2955 ± 220, 2170 ± 125, 1020 ± 115 BP) were taken, the oldest one of which was near the bottom of the core hole at c. 5.0 m. Nevertheless, the spread of the dated samples along the column, and the body of other proxy-data available, in addition to 518O/16O,13C/12C ratios (i.e., percentage of CaCO3) and the detailed pollen analysis (Baruch, 1986; Stiller et al., 1983-84), enable this time series to be used to interpret climate changes in the region during the upper half of the Holocene.

These data have been used by the author in his argument against the prevailing paradigm which claims that no significant climate changes occurred during the upper part of the Holocene and attributes all environmental changes to human activity (Issar, 1990). This is also the case with the data from the Sea of Galilee (Stiller et al., 1983-84), which were initially interpreted as reflections of anthropogenic factors rather than climate changes.

The examination of this core (Fig. 1.2) enables us to distinguish various zones. Zones of high 518O/16O and 513C/12C ratios are fromc. 5.0kato4.5ka, from2.8kato 2.3ka,fromc. 1.5kato c. 1.2ka and, finally, at 0.4 ka BP. Zones with only high <513C/12C ratios are from 4.2ka to 3.5 ka and at 1.8 ka BP. Toward the uppermost part of the 518O/16O curve, starting at c. 0.3 ka BP, there is a trend to heavier ratios.

The other 518O and 513C time series presented in Fig. 1.2 average the results of 41 stalagmites taken in 10 caves in Galilee,

Paleo-humidily Lake Van Levels of Mediterranean(m) Paleo-precipilalion (mm/year) Sorcq Cave

Based on data from Baaed on daUi rrom Pale o-levels Dead Sea Based on data from

Lemcke and Slurm(l997)* Raban & Galili(1985)* Frumkin et al. (1991) Bar-Mattlicws et al (1998a)*

Fig. 1.3. Paleo-hydrology time series in the Middle East. 1 Adjusted to scale and streamlined (3-5) points by the running average method.

northern Israel. The age determination for all the sequence was from calibration of 14C and uranium/thorium (234U/230Th) dates. Precisions of c. 300 years have been obtained, taking a reservoir effect of 900 years into account (Geyh et al., unpublished data). Therefore, speleotheme ages are considered to be calibrated dates with a ± 300 years margin of error. That these dates, within this margin of error, are reliable can be deduced from the close similarity between this speleotheme curve and the one from the Sea of Galilee. Periods of heavy isotopic composition occurred c. 4 ka, 3.8 ka and 1.2 ka BP, while periods of light compositions occurred c. 4.8ka, 3.3ka, 2.0ka and 1.0kaBP.

Another sequence of S18O/16O, forming a time series of paleo-climatic significance, is of a speleotheme from a cave in the vicinity of Jerusalem, in the mountainous part of central Israel (Fig. 1.2; Ayalon etal., 1998; Bar-Matthews etal., 1991, 1993, 1996, 1997, 1998a,b). The age determinations were made by the 230Th/234U method (Kaufman et al., 1998). The isotopic record, which is traceable for the last 58 ka, shows a pronounced difference between the values characterizing the speleothemes that were formed before 6.5 ka BP and those formed later, including the contemporary deposits. This, according to Bar-Matthews et al. (1998a,b), is probably because of altogether different climatic regimes.

This is an important observation with regard to the exact time dimension that is suitable to provide proxy-data for simulations using general circulation model (GCM) scenarios. Climate scenarios of the Pleistocene (glacials and interglacials) are not suitable whereas that starting c. 6ka BP is. With regard to the climate changes during the last 6.5 ka, the team working on the speleothemes of Soreq Cave (Ayalon et al., 1998; Bar Matthews et al., 1998a,b) have calculated the paleo-rainfall values by correlating the paleo-S18O records with the contemporary ratios of S18O/rainfall. Based on the S18O and S13C values and calculated paleo-rainfall, they divide the record into four stages. Stage 1 lasted from 6.5 ka to 5.4 ka BP and was very wet. During the period extending from 5.6ka to c. 3.0 ka BP (stage 2), the climate was, in general, humid, interrupted by four short dry spells. One was between 5.2ka and 5.0ka and another was at c. 4.0ka BP. Stage 3, lasting from c. 3.0ka to c. 1.0ka BP, was transitional to drier and more stable conditions. Stage 4, from c. 1.0 ka BP to the present, was characterized by fluctuations in rainfall. The high values between 0.4 ka and 0.5 ka BP may be connected with the Little Ice Age, while the increase in S13C values, which started c. 0.7 ka BP, may indicate a process of deforestation and increased grazing during the Turkish period.

The curve of S18O composition of pelagic and planktonic foraminifers (Luz, 1991) is not too conclusive. In general, changes in the S18O/16O values reflect changes in oceanic temperatures and the water in which the animals lived. Such ratios in foraminifers' shells in deep-sea sediments enabled the establishment of the sequence of climate changes during the Quaternary (Emiliani, 1955).

This, however, is mainly seen in planktonic assemblages. The iso-topic record for the benthonic forms shows low fluctuation because of the relatively stable temperature of the water at the bottom of the sea. Shackleton and Opdyke (1973) have demonstrated that the changes in isotopic values reflect the changes in the continental volume of ice as melted glacier water causes the water of the oceans to become isotopically lighter. It is certain that the fluctuations in the isotopic composition between glacial and interglacial periods resulted from the glacial effects (Bowen, 1991). However, it seems that the isotopic composition of the Mediterranean Sea is more complicated as the isotopic record from cores taken from the Mediterranean Sea also seems to reflect local changes. Consequently the isotopic composition of the Mediterranean Sea reflects not only climatic parameters such as precipitation, evaporation and residence time of water mass within the basin but also the hydrological regimes of the Black Sea, the Nile and the Atlantic Ocean.

From the ratio of S18O/16O of the epi-pelagic foraminifer Globigerinoides rubes, Rossignol-Strick et al. (1982) found that the oxygen isotopic composition ratio decreased through several large shifts to minimal values between 8 and 6 ka BP (see also Luz and Perelis-Grossowicz, 1980). The sharp depletion in 18O/16O ratio and the lowering in salinity from 8 ka to 7 ka BP may represent the heavy Nile floods during a mainly rainy period in Africa (Nicholson, 1980; Nicholson and Flohn, 1980), and it seems likely that the Nile was the major source of fresh water responsible for the low salinities in the Mediterranean Sea. However, the Nile water, coming from areas of low latitude, should be isotopically heavy. In a recent work, Luz (1991) suggested an alternative explanation for the isotopic depletion. He claimed that a high influx of low-salinity water entered the Mediterranean Sea from the Black Sea when the rising sea surface reached the level of the Bosphorus. This alternative explanation is not in agreement with the findings of Erinc (1978), who concluded that, even though the sea level rise in the Black Sea and in the Mediterranean Sea started simultaneously after the peak of the last glacial period, the Black Sea basin was disconnected from the Mediterranean. Moreover, the rise in sea levels caused the intrusion of the Mediterranean into the Black Sea. Cores obtained in the Black Sea (Degens, 1971; cited in Erinc, 1978) indicate, "that the main intrusion of saline water into the Black Sea started 7140 ± 180 years ago". In conclusion, it is clear that the reasons for the changes in oxygen isotope composition of the foraminifers of the Mediterranean have yet to be elucidated. In their book Noah's Flood: The New Scientific Discoveries about the Event that Changed History, the marine geologists William Ryan and Walter Pitman (1998) argue that this intrusion (around 7500 years ago) was caused by the breaching of the barrier at the Bosphorus, at the northeastern part of the Sea of Marmara, which filled up an ancient lake, the predecessor of the Black Sea, the level of which was 150 mlower than the present sea level. This caused a tremendous waterfall of seawater flooding the lowlands surrounding the ancient lake, a calamity for the people in the Neolithic agricultural communities that lived in this region. They further claim that this calamity lived on in the memory of the people who survived and migrated into Mesopotamia. The stories told were passed on from one generation to the next until they crystallized in the mythological texts found on ancient clay tablets of ancient Mesopotamia. At a later period, these stories were incorporated into the Hebrews' sacred scriptures and became part of the Judeo-Christian-Moslem heritage. Issar and Zohar (2003) maintain that the findings of an ancient flood filling up the Black Sea to its brim is, undoubtedly, of the greatest importance for the understanding of the prehistory of Europe and Central Asia during the Lower Holocene (i.e., 10 ka to 5 ka BP). And yet, this discovery should not be mixed up and confused with the Biblical Flood.

Schilman et al. (2001, 2002) examined two cores drilled at the sea bottom in the southeastern part of the Mediterranean near the shores of Israel for oxygen and carbon isotope composition as well as for physical and geochemical properties of the sediments. The date of the lowest layer of the sequence is c. 3.6 ka BP. According to Schilman et al. (2001, 2002), the S18O values of the planktonic foraminifer G. ruber suggest that humid phases took place between 3.5 ka and 3.0 ka and between 1.7 ka and 1.0 ka BP, while arid conditions prevailed between 3.0 ka and 1.7 ka BP. At c. 0.8 ka BP, a warm period, the Medieval Warm Period, took place and at 0.27 ka BP, a cold period, the Little Ice Age, occurred. Schilman et al. (2001, 2002) also suggest a long-term trend of aridization that started c. 7.0 ka BP in the mid-low-latitude desert belt and has continued until the present. They base their suggestion on the long-term slight increase in S18 O values of planktonic foraminifers, which corresponds with a gradual decrease in the S13C values of both G. ruber and the benthos foraminifers Uvigerina mediterranea. This trend is concurrent with an increase in sedimentation rates, the titanium/aluminum (Ti/Al) ratio, magnetic susceptibility and color index of the sediments. Schilman et al., suggest that this general long-term warming up, and thus aridization, reflects a gradual change in the S13C of the dissolved CO2 of the entire southeastern Mediterranean water column, which parallels the global rise of atmospheric CO2 observed for the late Holocene. They suggest that this is a result of terrestrial biomass destruction during the aridization process and the gradual reduction of the vegetation cover in east Africa, which led "to an increased erratic flood-related sediment flux via the Nile River. This is reflected by the general change in the local sediment composition. At 3.6 ka ago, the Saharan eolian input reached 65% whereas at about 0.3 ka ago 70% of the SE Mediterranean sediment was composed of Nile particulate-matter." I prefer to put more emphasis on the relative fluctuations of the isotopic composition and sedimentary sequence rather than on the general trend.

As can be seen from Fig. 1.2, the two peaks of light oxygen isotopes (c. 3.40 ka and 1.4 ka BP), which suggest an influx of melted glacial water (i.e., a warm climate), correspond to two major warm periods: the Late Bronze and the Arab period. A secondary light oxygen period occurred at c. 0.7 ka BP, which corresponds to the Mamluk-Ottoman warm phase. The two peaks of heavy oxygen isotopes (c. 2.30 ka and 0.3 ka BP) correspond to two cold periods: the Roman and the Little Ice Age. A secondary cold period occurred at c. 1.0 ka BP, which corresponded to the Crusader period. I would also interpret the change in the sediment characteristics differently. The higher loess load at c. 3.60 ka BP could be the result of the inflow of loess from the higher rate of dust storms and floods in northern Egypt, Sinai and Negev during the Middle Bronze Age (MB), which was relatively (to the Intermediary and Late Bronze periods) cold and humid. The higher sand supply at c. 0.3 ka BP would be a function of the general warming up that started at c. 1.4 ka BP, which brought higher rates of easterly rainstorms over northeastern Africa and higher supplies of sand from the Nile. This corresponds with the post-Byzantine invasion of sand dunes into the coastal plain of Israel.

The curve reconstructing the sea-level changes along the coastline of central Israel, presented in Fig. 1.3, is by Raban and Galili (1985). It is based on a survey of archaeological sites along the Israeli coastline, with submarine as well as surface structures. It incorporates the results of the work of Galili et al. (1988), who reconstructed ancient sea levels between 8 ka and 1.5 ka BP along the coast line of Mount Carmel, and the conclusions of the survey of Bloch (1976), who based his observations on the altitude of ancient salt production basins. The curve of Raban and Galili (1985) shows that, since the lower Holocene, the sea level has risen to reach that of the present day. Between 8 ka and 6 ka BP, the sea level rose at a mean annual rate of 5.2mm. According to these authors, no tectonic movements have occurred in the area during the last 8000 years. The most pronounced recessions of sea levels shown on this curve are between 4.5 ka and 4.0 ka BP, between 3.5 ka and 3 ka BP, between 2.5 ka and 2ka BP and c. 0.7 ka AD. The periods of high sea level are around 5 ka BP, between 4 ka and 3.5 ka BP, from 3 ka to 2.5 ka BP and c. 1.4 ka BP. A trend toward a higher sea level can be seen after 0.5 ka BP. It is suggested that the periods of low sea levels correlate with periods of cold climate, that is, periods of expansion of polar glaciers, while during periods of high sea level the climate was warm and the glaciers melted.

While high levels of the Mediterranean indicate periods of warm climate and vice versa, high levels of the Dead Sea during the Holocene indicate cold and humid periods. This is because the Dead Sea is located at the lower end of the Jordan catchment basin, and its levels are determined by the amount of precipitation on this basin and the rate of evaporation from its surface. The curves of the levels of the lake presented in Fig. 1.3 are based on a survey of ancient shorelines and erosion channels inside the salt caves of Mount Sodom (Frumkin et al., 1991). The results are in agreement with a prior survey that was based on ancient shorelines (Klein, 1982). The periods of high lake levels were found to have occurred c. 8 ka, 4.5 ka, 3 ka, 2 ka and 1 ka BP, while periods of very low levels, which most probably caused the drying up of the southern part of the Dead Sea, occurred c. 6.5 ka and 2.5 ka BP. In assessing the evidence from these curves, we have to take into consideration the fact that higher and younger lake levels may obliterate the evidence of older but lower levels.

A rather similar pattern of climate changes could be deducted from the carbon and oxygen isotope values investigated in a speleotheme in Nahal Qanah Cave, central Israel (Frumkin et al., 1999).

0 0

Post a comment