EARLY HOLOCENE

Chapter 1 (Pages 25-31)

Pre-History and Archaeology Glossary

Excerpts and Definitions and Addendums:

The date for the end of the Pleistocene and beginning of the Holocene in northern Europe has been determined as about 8000 BC, one recent estimate being 8300 BC (See Page 530 in *1 Below). This date is believed to mark a substantial reduction of the glaciers and the draining of the Baltic Lake. These criteria have little relevance for the Levant yet this date does appear to coincide with certain environmental changes there that characterise the true Holocene. 8000 B.C. may therefore also be taken as a convenient point at which to divide the Pleistocene and Holocene in the Levant.

After a marked fall in the 9th millennium BC the temperature rose again at the beginning of the Holocene and continued to increase until perhaps as late as 3000 BC, certainly throughout the Neolithic. This rise seems to have been less irregular in the earlier Holocene than during the Late Glacial but there may still have been minor fluctuations that affected the environment of the Levant.

The Dead Sea remained at a higher level than before for several millennia during the early Holocene. The ratio of run-off to evaporation was high during this period and an extensive bed of clay was laid down in the Dead Sea basin (See Page 28 in *2 Below). During a subsequent drier phase a layer of rock salt was superimposed on this clay bed. The deposition of this rock salt is estimated to have taken place between 4500 and 3500 BC. Another phase of increased available moisture followed when a further bed of clay was laid down; this bed is dated by a single C-14 determination of 2460 ± 320 B (ibid). The Beth Shan lake existed throughout this period but was finally drained sometime after 3000 BC.

No significant alterations in lake level during the Holocene have been detected in the Jafr depression but changes were taking place in the Damascus basin. Here alternating drier and moister phases are believed to have influenced the environment during the Holocene although these phases are not well dated. During the middle Holocene a deep layer of calcareous sediments was deposited by heavy winter rains at Tekieh in the Barada gorge and run-off was sufficiently great to wash more sediment into the Damascus basin (See Page 54 in *3 Below). Nevertheless, despite this evidence of erosive activity, the late Wurm lake in the Damascus basin gradually shrank and divided into two to form the present Hijjane and Ataibe lakes. This was caused by an increase in evaporation following the Holocene rise in temperature and an eventual decrease in effective run-off.

The rise in sea-level towards the end of the Pleistocene sharply reduced the gradient of the rivers on the seaward side of the Jebel Alawiye. The valleys of the Nahr el Kebir and the other smaller rivers began to fill with debris (See Page 24 ibid). The erosive power of the run-off in this region has not been very strong during the Holocene so this deposit has continued to accumulate and little sediment reaches the sea.

The evidence for changes in vegetation in the Levant during the Holocene is less satisfactory than for the Pleistocene. Very few pollen cores have been drilled and studied while those which are available are difficult to interpret. This is true even in the northern Jordan valley where no less than five cores have been bored in Holocene sediments, three in the Huleh basin (K-Jam - U.P.6 - U.P.15) and two in the bed of the Sea of Galilee (D-10l6/2 and D-1021/1). Unfortunately the cores were sampled at rather wide intervals and relatively few pollen grains were recovered (See Page 259 in *4 Below). C-14 determinations have been made on samples from only three of the five cores: K-Jam - U.P.15 - D-1016/2. The K-Jam date (Hv-1725) falls in the Pleistocene but the other dates cluster in the second half of the Holocene. Because of this uneven distribution one must treat with caution Horowitz's calculations of average sedimentation rates in the core particularly when he applies the supposed rate in the dated core U-P-15 to U.P.6 although the lithology in the two is completely different (See Page 269 ibid). One's misgivings are reinforced on observing that some of the dates applied to different stages in the cores have been incorrectly calculated from the supposed sedimentation rates (See Page 268 ibid). Even when one allows for these discrenancies one finds that on Horowitz's own division of the stages the results from the cores are in part contradictory.

In spite of these difficulties it is necessary to attempt to analyse the data presented in this study because it provides almost the only independent evidence about the Holocene vegetation in this region; the other source of data, evidence from archaeological sites, may be expected to be influenced by man's activities. K-Jam is a much deeper core than the others and it is only the section from 18 to 9 metres which is believed to record the vegetation in the first half of the Holocene (See Page 267 ibid). At 18 metres the percentage of arboreal pollen, principally oak, is relatively high but then this gradually declines. At the same time pollen of marsh plants and grasses increases, reflecting a gradual contraction of Lake Huleh.

A more detailed picture of Holocene vegetation changes may be obtained from core D-1016/2. This core is thought to have been drilled entirely in Holocene sediments (See Page 270ff ibid) and samples for pollen analysis have been taken at shorter intervals of time than in the K-Jam core. A sample at the bottom of the core from a depth of 30 metres below the surface of the lake which is believed to date from the beginning of the Holocene has a low value of 10% for arboreal pollen composed mostly of oak with a little Aleppo pine. Open field species are much more common, accounting for over 55% of the pollen (See Diagram on Page 266 ibid). Horowitz believes that these figures are supported by the pollen spectra of the other undated core from the Sea of Galilee (D-1021/1) [See Table 16 ibid]. These results directly contradict the evidence from the K-Jam cores which suggested that Galilee was well-forested at the beginning of the Holocene.

Subsequently the values for arboreal pollen increase in core D-1016/2, reaching a peak between 27 and 23 metres, a little before the middle of the Holocene. The proportions of oak pollen in the spectra have actually declined but walnut and jujube make up the diffcrence. Open field pollen still forms an important percentage of the total. Again it would appear difficult to reconcile this curve for arboreal pollen with that from the K-Jam core.

The two other cores from the Huleh basin (U.P.6 and U.P.15) are believed to extend back only as far as the earlier Holocene (See Page 268ff ibid). The percentages of arboreal pollen are quite high at the bottom of these cores, possibly correlating with the relatively high values we have noted for the earlier Holocene in core D-1016/2. The pollen is mostly from oak trees with a little pistachio and olive but no jujube; the high values for the latter species in the D-1016/2 core may be anomalous.

Can the often apparently conflicting evidence from these cores be brought together to present a coherent picture of vegetation change? Part of the difficulty may be that the record from the K-Jam core is a long one and insensitive to short-term variations. The evidence from this core suggests a steady decline in forest vegetation as the Holocene progresses, a general trend apparent from the arboreal pollen curves in the two other Huleh cores, though not fully supported by the evidence from the cores from the Sea of Galilee. The low arboreal pollen values for the beginning of the Holocene in core D-1016/2 may refer to a short-term fluctuation in that area which has not been noticed in the more generalised studies of the K-Jam core.

One may now present with due caution a synthesis of the vegetation record from these cores. At the close of the Pleistocene when the Huleh lake was quite extensive the hills of Galilee and the Golan were clothed with forest composed principally of oak. At the beginning of the Holocene there may have been a brief decline in forest cover, at least in the environs of the Sea of Galilee, but thereafter the hills were forested until the mid-Holocene. This long period of fairly extensive tree cover in northern Palestine coincides with the phase of greater run-off that maintained the Dead Sea at a high level. Later in the Holocene most of the cores show a decline in tree pollen and in the Huleh basin an extension of marsh vegetation which Horowitz believes can be attributed to generally drier conditions (See Page 268 ibid). It would seem more likely by this time that man was partly responsible for these vegetation changes through forest clearance.

If this interpretation is correct for Galilee and the Golan then one may tentatively infer from it something of the pattern of vegetation further south. It is likely that the Judean hills and the mountains of Transjordan were forested at least as far south as the latitude of the Dead Sea and probably further. Timber from a number of species from the forested zone such as ash, plane, almond, pear, olive, fig, Christ's Thorn and carob was used by the Neolithic inhabitants of Jericho (See Figure 1 in *5 Below). This suggests that the Mediterranean forest lay nearer the site than it does today (See Page 40 ibid).

Northern Sinai would have been affected by the moister conditions that existed in Egypt and so probably had a richer vegetation in the earlier Holocene than now. This was almost certainly the same vegetation of steppe and scattered trees that existed in the area at the end of the Pleistocene. Only when more arid conditions returned after 6000 BC would the vegetation have deteriorated, a gradual process that has continued until the present day. Even now a few relict specimens of trees are to be found in favoured habitats in both Sinai and Transjordan (See Page 54 in *6 Below) though the vegetation zones of which they were characterisic have retreated.

There is very little eviaence from which to determine the vegetation in Lebanon and southern Syria during the Holocene. Presumably the Mountains of Lebanon remained heavily fcrested for several thousand years. A single pollen spectrum from Tekieh suggests that the Anti-Lebanon still carried some trees and herbaceous ground cover as late as the mid-Holocene. Further east steppe vegetation characterised by Chenopodiaceae and Compositae began to encroach upon the Damascus basin and the Golan quite early in the Holocene but it is likely that man's activities hastened this transformation.

The principal source of information for the vegetation of northern Syria during the earlier Holocene is the Ghab pollen core. Sub-zone Z3 in the pollen diagram is ascribed to this period (See Page 750 in *7 Below). Jebels Alawiye and Zawiye remained clothed with light mixed forest throughout this period. The forest was principally composed of evergreen oak, olive, pine (Pinus brutia), pistachio and hornbeam with some cedar of Lebanon. There was less oak pollen in this sub-zone than in Z2 and a higher percentage of herbaceous pollen. This may indicate that the tree cover was less dense than in Z2 and that rainfall decreased slightly (ibid) in the early Holocene. Almost the same type of light forest would be the natural vegetation cover in this area today were it not for man's interference. Later in the Holocene species such as pistachio, olive, juniper and hornbeam decreased while cedar disappeared completely because of man's activities (See Page 751 ibid).

The vegetation around Abu Hureyra during the earlier Holocene was steppic but still richer in species than today (See Page 70 in *8 Below). This would indicate that the total rainfall was probably about the same as now although it may have been more regular. The drastic degradation of the steppe that occurred during the Holocene around Abu Hureyra and throughout northern and central Syria was brought about more by man's activities than by climatic change.

It will be useful now to compare the data we have already considered with recent proposed reconstructions of post-glacial atmospheric circulation in the northern hemisphere. During the 9th millennium atmospheric circulation is believed to have been relatively weak over Europe and it is thought that the belt of westerly winds lay further south than today (See Page 140ff in *10 Below). Rain-bearing winds from the Atlantic and Mediterranean would have reached the Levant more frequently (See Figures 10b and 10c in *9) and so winter rainfall would have been greater than today; it is also possible that the rainy season lasted longer than now.

By about 6500 BC the European ice sheet had almost disappeared but the North American one still existed. This stimulated a more vigorous wind system and also caused the westerlies over the Atlantic to follow a more northeasterly course. A strong anticyclone lay in the eastern Atlantic and over western Europe (See Page 160 and Figures 8 and 9 in *10 Below), preventing some of the moist air from the Atlantic reaching the Levant. In consequence rainfall became more seasonal and possibly decreased in amount; the continued post-glacial rise in temperature would have also tended to lower precipitation.

About 4000 BC the North American ice sheet had melted so the wind system resumed a more zonal configuration but now the track of the westerlies lay further north (See Page 160 and Figure 10 in *10 and Figure 11g in *11). The Levantine summer became longer and drier while fewer rain-bearing depressions penetrated from the west in winter. Consequently this was a relatively arid period. Subsequently the atmospheric circulation fluctuated again (See Page 160ff in *10 Below) and this led to an amelioration of climatic conditions in the Levant during the following millennia ...