Millennial multi-proxy reconstruction of oasis dynamics in Jordan, by the Dead Sea

Abstract

Vegetation reconstructions in the Dead Sea region based on sediment records are potentially biased, because the vast majority of them derive from the western side of the sea, and only focus on large areas and time spans, while little is known about extra-local (< 1,000 m radius) to local (< 20 m radius) changes. To fill this gap, we compared a vegetation survey with modern pollen assemblages from the “Palm Terrace” oasis ca. 300 m b.s.l. (below sea level), at the eastern edge of the Dead Sea. This revealed how the oasis vegetation is reflected in pollen assemblages. In addition, two sediment cores were collected from the centre and the edge of a mire at the oasis to reconstruct past vegetation dynamics. We analysed sedimentary pollen and microscopic charcoal, as well as the sediment chemistry by X-ray fluorescence (XRF) and conductivity, focusing on the past ~ 1,000 years. Pollen results suggest that mesophilous Phoenix dactylifera (date palm) stands and wetland vegetation expanded there around ad 1300–1500 and 1700–1900. During the past ca. 100 years, drought-adapted Chenopodiaceae gained ground, partly replacing the palms. Results from elemental analysis, especially of elements such as chlorine, provide evidence of enhanced evaporative salinization. Increasing desertification and the associated decline of mesophilous date palm stands during the past ca. 50 years is probably related to a decrease in annual precipitation and also corresponds to decreasing water levels in the Dead Sea. These have mainly been caused by increasing extraction of fresh water from tributaries and wells, mainly for local agriculture and industry. In the future, with hotter and drier conditions as well as increased use of water, oasis vegetation along the Dead Sea might be at further risk of contraction or even extinction.

Full text

Vol.:(0123456789)
1 3
Vegetation History and Archaeobotany
https://doi.org/10.1007/s00334-017-0663-6
ORIGINAL ARTICLE
Millennial multi-proxy reconstruction ofoasis dynamics inJordan,
bytheDead Sea
SebastianEggenberger1· ErikaGobet1· JacquelineF.N.vanLeeuwen1· ChristophSchwörer1 ·
WillemO.vanderKnaap1· HanF.vanDobben2· HendrikVogel3· WillyTinner1· ClaireM.C.Rambeau1,4
Received: 4 April 2017 / Accepted: 7 December 2017
© Springer-Verlag GmbH Germany, part of Springer Nature 2017
Abstract
Vegetation reconstructions in the Dead Sea region based on sediment records are potentially biased, because the vast major-
ity of them derive from the western side of the sea, and only focus on large areas and time spans, while little is known about
extra-local (< 1,000m radius) to local (< 20m radius) changes. To fill this gap, we compared a vegetation survey with modern
pollen assemblages from the “Palm Terrace” oasis ca. 300m b.s.l. (below sea level), at the eastern edge of the Dead Sea.
This revealed how the oasis vegetation is reflected in pollen assemblages. In addition, two sediment cores were collected
from the centre and the edge of a mire at the oasis to reconstruct past vegetation dynamics. We analysed sedimentary pollen
and microscopic charcoal, as well as the sediment chemistry by X-ray fluorescence (XRF) and conductivity, focusing on the
past ~ 1,000years. Pollen results suggest that mesophilous Phoenix dactylifera (date palm) stands and wetland vegetation
expanded there around ad 1300–1500 and 1700–1900. During the past ca. 100 years, drought-adapted Chenopodiaceae gained
ground, partly replacing the palms. Results from elemental analysis, especially of elements such as chlorine, provide evidence
of enhanced evaporative salinization. Increasing desertification and the associated decline of mesophilous date palm stands
during the past ca. 50years is probably related to a decrease in annual precipitation and also corresponds to decreasing water
levels in the Dead Sea. These have mainly been caused by increasing extraction of fresh water from tributaries and wells,
mainly for local agriculture and industry. In the future, with hotter and drier conditions as well as increased use of water,
oasis vegetation along the Dead Sea might be at further risk of contraction or even extinction.
Keywords Pollen· Fire history· Global change· Vegetation· XRF· Phoenix dactylifera
Introduction
The Dead Sea basin is situated in the Dead Sea Transform
(DST), a series of faults that shape the geology and geomor-
phology of the area (Garfunkel and Ben-Avraham 1996).
With a water level of 430m b.s.l. (below sea level) in 2015
(Yechieli etal. 2016), the Dead Sea is the lowest-lying water
body on Earth. Its unique location combined with an extreme
climate and ancient cultural legacies give special impor-
tance to this region. Whereas the whole Arabian Peninsula
currently suffers from water scarcity due to the increasing
water demands of growing populations (Odhiambo 2017), in
Jordan this trend has been aggravated by massive immigra-
tion waves from neighbouring countries such as Syria and
Iraq (Chen and Weisbrod 2016). It is therefore crucial to
understand the various processes coupled to the hydrologi-
cal cycles that may affect land use and vegetation dynamics.
Communicated by T. Litt.
Electronic supplementary material The online version of this
article (https://doi.org/10.1007/s00334-017-0663-6) contains
supplementary material, which is available to authorized users.
* Christoph Schwörer
christoph.schwoerer@ips.unibe.ch
1 Institute ofPlant Sciences andOeschger Centre forClimate
Change Research, University ofBern, Altenbergrain 21,
3013Bern, Switzerland
2 Wageningen University andResearch, PO Box47,
6700AAWageningen, TheNetherlands
3 Institute ofGeological Sciences andOeschger Centre
forClimate Change Research, University ofBern,
Baltzerstrasse 1+3, 3012Bern, Switzerland
4 Institute ofEarth andEnvironmental Sciences, University
ofFreiburg, Albertstraße 23b, 79104FreiburgimBreisgau,
Germany

Vegetation History and Archaeobotany
1 3
In the Dead Sea region, available palynological studies
mainly focus on the western, Israeli side (Baruch 1993;
Heim etal. 1997; Neumann etal. 2007, 2009, 2010; Leroy
2010; Leroy etal. 2010; Litt etal. 2012; Langgut etal.
2014, 2015), whereas research on the eastern, Jordanian,
side is scarce (Rambeau 2010). The vegetation at the edge
of the Dead Sea is characterized by striking differences
in composition within small distances (Davies and Fall
2001). Our comparison of surface pollen with vegeta-
tion surveys of a small (~ 2ha) oasis, in this study called
“Palm Terrace”, allows us to investigate how extra-local
(< 1,000 m radius) to local (< 20m radius; Jacobson
and Bradshaw 1981) vegetation patterns are reflected in
recent pollen assemblages. This comparison is used to
refine the palynological vegetation reconstructions from
two sediment cores from “Palm Terrace”, core PT1 from
the central part and core PT2 from the edge of today’s
wetland. In order to gain a better understanding of past
environments, sedimentological proxy evidence from
X-ray fluorescence (XRF) and conductivity measure-
ments was included. The aim of this study is to provide
new insights into the palaeo-environmental dynamics at
the oasis over the past approximately 1,000years. This
time period is under-investigated in the region, since most
published studies do not reach to the present or are too
coarse in time resolution for detailed interpretation. We
briefly discuss the implications of our study for global
change biology and the future of desert oases.
Study site
“Palm Terrace” (PT) is a small ~ 2ha oasis located in north-
western Jordan (31°32′14″N, 35°33′45″E, 290–335m b.s.l.)
in the governorate of Madaba (Fig.1). It is situated 700m
east of the Dead Sea shore (northern basin), 7km south of
the hot springs of ‘Ain ez-Zara’ (ancient Greek: Kallirrhoë),
and 8km north of Wadi Mujib. The site is part of the Mujib
Biosphere Reserve, the lowest-lying nature reserve on Earth,
established in 1985 by the Jordanian Royal Society for the
Conservation of Nature. Sources from Roman historians and
rabbinical traditions, as well as archaeological findings from
Kallirrhoë, prove human activities in the close surround-
ings of the oasis during the centuries around the year ad, as
well as during the early Byzantine period in the fourth and
fifth century ad (Clamer 2010). On the famous mosaic map
from Madaba dating from the sixth century ad, Kallirrhoë
is labelled ΘΕΡΜΑΚΑΛΛΙ ΡΟΗC and Wadi Mujib can be
recognised (Hirschfeld 2006). The two palms depicted in
between might represent the “Palm Terrace” oasis (Fig.2).
Tectonically, the oasis is situated on a fault parallel to the
Dead Sea Transform (DST) (Garfunkel and Ben-Avraham
1996), and traces of faulting can be observed on the escarp-
ment. This rises on the eastern side of the oasis as a steep
slope of sandstone that reaches over 1,000m a.s.l., where it
merges into the Madaba and Dhiban plateaus. Whereas the
spring from ‘Ain ez-Zara’ is fed by water from an aquifer
complex in the western highlands 20–30km further east
(Salameh and Udluft 1985; Rimawi and Salameh 1988;
Sawarieh etal. 2004, 2009), it is not yet clear where the
Fig. 1 Left: Topographical map of the southern Levant showing the
Dead Sea and the location of the study site “Palm Terrace”. Inset
shows the location of the study site in the Middle East. Right, map
showing the vegetation zones in the area (after Zohary 1962; Langgut
et al. 2015) and important palaeoclimatological and archaeological
sites in the region

Vegetation History and Archaeobotany
1 3
water of “Palm Terrace” originates from. Today’s climate
is characterized by extremely dry and hot conditions. Aver-
age annual rainfall over the Dead Sea is ~ 90mm, most of
which falls between October and May (Dayan and Morin
2006). The mean annual temperature is 25.9°C, the hottest
month being August with an average of 32.9°C, the coldest
January with 18.3°C (Hect and Gertman 2003). Steep slopes
combined with harsh climatic conditions make the eastern
Dead Sea coast unfavourable for inhabitation. As a result of
the topography, climatic gradients are also steep, creating
strong environmental gradients and clear associated vegeta-
tion belts (Fig.1, Zohary 1962; Davies and Fall 2001; Neu-
mann etal. 2010; Langgut etal. 2015). The highest belt, the
Mediterranean zone, occurs about 20km east of the Dead
Sea on the Madaba and Dhiban plateaus and is character-
ized by evergreen maquis scrub with Quercus calliprinos
(evergreen oak) and Juniperus phoenicea. Below it follows
the Irano-Turanian vegetation zone on the upper slopes
towards the Dead Sea, dominated by steppe grasslands with
Artemisia herba-alba and other plants. Below this belt, the
Saharo-Arabian vegetation zone with salt-tolerant Chenopo-
diaceae and Tamarix spp. grows in desert conditions. Inside
this belt are azonal patches of Sudano-Deccanian vegeta-
tion, linked to freshwater springs (Zohary 1962; Davies and
Fall 2001; Albert etal. 2004; Neumann etal. 2010; Langgut
etal. 2015). “Palm Terrace” is such an example of azonal
vegetation distribution. While situated in the Saharo-Arabian
zone, it also includes wetland-adapted plants belonging to
Sudano-Deccanian vegetation such as Phoenix dactylifera,
and to Mediterranean and Irano-Turanian vegetation such as
Typha domingensis and Saccharum ravennae.
Materials andmethods
Vegetation mapping ofthestudy site, surface
pollen samples
To reconstruct the vegetation history and to investigate the
long-term response of plant communities to past climatic
changes and disturbances, it is crucial to understand the
linkage between today’s vegetation and the palynological
signal that it produces. To our knowledge, no data exist
about the palynological representation of oasis vegetation
in the Jordan valley, although vegetation zones along an alti-
tudinal transect could be distinguished by modern pollen
precipitation (Davies and Fall 2001). In February 2016, we
determined plant occurrence and abundance in 123 relevés
(survey areas) of ~ 100m2 (Braun-Blanquet 1964) and col-
lected six surface soil samples that were later analysed for
pollen composition. The vegetation results were mapped
using a Garmin eTrex Summit GPS receiver with an accu-
racy of ± 10m and summarized on a satellite image (Google
Earth 2004; Fig.3; ESM Fig.1). The elevation in m b.s.l.
was calculated from a local coordinate system derived from
total station surveying, which was carried out by a Jorda-
nian topographical survey team, surveyor Yehya Suleiman
Al-Hasanat.
Coring andsampling
In March 2015, the 197cm long sediment core PT1 from
the centre of the “Palm Terrace” mire (31°32′14.78″N,
35°33′44.13″E) was taken with a Russian peat corer with
a 5cm diameter. To support the results from this peat core
and to better understand past oasis dynamics, the 246cm
long sediment core PT2 was retrieved in February 2016
a few metres outside the active mire (31°32′16.06″N,
35°33′45.86″E) using the same coring technique. The dis-
tance between the cores is ~ 65m. Both cores were sub-
sampled for pollen, microscopic charcoal and macroscopic
plant remains. Core PT1 provided terrestrial plant material
suitable for radiocarbon dating and 34 subsamples for pol-
len and microscopic charcoal analyses (1cm3; at 186 and
170cm: 2cm3) were taken between 194 and 7cm depth with
an average distance of ~ 6cm between samples (Table1).
From core PT2, 10 subsamples (1cm3) were taken between
240 and 30cm depth.
Fig. 2 Detail of mosaic map from the sixth century ad in St. Georges
Church in Madaba, Jordan. Note the broad river Jordan (1) enter-
ing the Dead Sea from the north. On the east shore of the Dead Sea,
south of Kallirrhoë (2; ΘΕΡΜΑΚΑΛΛΙ ΡΟΗC) and north of Wadi
Mujib (3), are two palms that might representour study site “Palm
Terrace” (4)

Vegetation History and Archaeobotany
1 3
Chronology
Three samples of macroscopic charcoal (each composed of
several particles > 200µm) and a rhizome from PT1 were
dated by accelerated mass spectrometry (AMS) 14C analy-
sis in the LARA AMS laboratory of the University of Bern
(Szidat etal. 2014, Table1; Fig.4). For calibration of the
14C dates, the IntCal13 Northern Hemisphere atmospheric
radiocarbon calibration curve (Reimer etal. 2013) and the
post-bomb curve for the Northern Hemisphere Region 2
(Hua etal. 2013) were used. The date of the rhizome was
not used for any of the calculations of the age-depth model,
Fig. 3 Map of “Palm Terrace” showing vegetation surveys, surface
pollen samples and conductivity. Red circles indicate the coring sites
PT1 and PT2. Magenta circles show vegetation plots from the outer
limits of present active mire, purple circles vegetation plots on a tran-
sect through the mire. Dark blue squares show the vegetation plots
from the outer limits of the Nitraria belt, light blue squares the addi-
tional plots in the Nitraria belt. Orange diamonds show vegetation
plots from the outer limits of a Phragmites zone. Green triangles rep-
resent single palm trees (Phoenix dactylifera). The blue spiral shows
the location of an active spring. The yellow star indicates a small
zone that was heavily influenced by grazing. Large symbols indicate
the position of surface pollen percentage diagrams [selected types:
Phoenix dactylifera (Pho), Poaceae (Poa), Typha domingensis-type
(Typh), Nitraria retusa (Nit), Tamarix nilotica (Tam), Chenopodia-
ceae/Amaranthaceae (Che)], the white bars show ×10 exaggerations.
The conductivity diagram shows the total dissolved solids (TDS) per
1cm3 on a logarithmic scale from surface samples taken either inside
the active mire or inside the Nitraria belt. Vertical thin lines indicate
the topography (m b.s.l.)
Table 1 Radiocarbon dates and
calibrated ages from core PT1
Calibration according to the IntCal13 Northern Hemisphere atmospheric radiocarbon calibration curve
(Reimer etal. 2013)
a Following Stuiver and Polach (1977)
b Postbomb for Northern Hemisphere Region 2 (Hua etal. 2013)
Lab. code Depth (cm) Material 14C Age(uncal BPa) Age 2σ range (cal ad) Age in
diagram
(cal ad)
BE-2647.1.1 60–64 Rhizome Modern Rejected –
BE-5610.1.1 65.5 Charcoal 225 ± 50 1516–1955b1628
BE-2648.1.1 93.5 Charcoal 695 ± 40 1255–1392 1388
BE-2649.1.1 151–153.5 Charcoal 770 ± 70 1047–1390 1164

Vegetation History and Archaeobotany
1 3
because we assume that it had penetrated into older layers,
which would explain its modern age. The smooth-spline
curve of the age-depth model (Fig.4, smooth-spline 0.3) was
computed with the program clam 2.2 (Blaauw 2010); the
95% confidence envelope was calculated according to gen-
eralized mixed-effect regression (Birks and Heegaard 2003).
Palynological analysis, presentation ofresults
andzonation ofthepollen diagram
Abundant pyrite framboids occurred in the sediments, espe-
cially in the deeper parts of the two cores, which made it
very hard to extract pollen with standard laboratory meth-
ods. We therefore applied two different pollen preparation
methods depending on pyrite content (Rambeau etal. 2015).
Samples with low pyrite content and the surface samples
were treated chemically with the standard method using
HCl, KOH, HF, sieving over 0.5mm, decanting, acetolysis
and storage in glycerine (Moore etal. 1991). Samples with
high pyrite content (PT1: 194, 187, 178, 171, 162, 154, 145,
138, 130, 126, 122cm; PT2: 228, 216cm) received a dif-
ferent treatment with HNO3 instead of acetolysis (Rambeau
etal. 2015). To calculate particle concentrations (particles
cm−3) and influx (particles cm−2 year−1), Lycopodium tab-
lets with a known number of spores were added prior to
processing (Stockmarr 1971). The pollen reference collec-
tion of the University of Bern together with pollen keys and
atlases (Reille 1992, 1995, 1998; Moore etal. 1991; Beug
2004; Phillips etal. 2013) were used for identifying pollen
and spores under a light microscope at 400 × magnification.
Because of extremely low pollen and spore abundances and
the very poor preservation towards the deeper ends of the
two cores, the minimum pollen sum was only 51 in PT1,
while no pollen was detected in the deepest samples of core
PT2. The average pollen sum in PT1 was 148. In the remain-
ing samples of PT2, the pollen sum varied between 57 and
704, with an average of 170 pollen grains per sample. In
the six surface samples, an average of 347 pollen grains
was counted per sample. Since wetland plants such as T.
domingensis, Phragmites australis and other members of the
Poaceae are the dominant components of the oasis vegeta-
tion and the main contributors to the pollen signal, their pol-
len was included in the pollen sum, whereas algae and fun-
gal spores were excluded. Chenopodiaceae/Amaranthaceae
(Cheno/Ams) were put in the herb sum, although this taxon
includes some shrubs. The pollen diagrams were drawn with
Tilia 2.0.41. Local pollen assemblage zones (LPAZ) were
identified by optimal sum of squares partitioning (Birks and
Gordon 1985) using ZONE v.1.2 (Juggins 1991), and their
statistical significance was tested with BSTICK (Bennett
1996). No statistically significant LPAZs were found in PT1
when the wetland plants were included. To find out if upland
vegetation history could be subdivided into statistically sig-
nificant LPAZs, we excluded the wetland plants from the
pollen sum and repeated the zonation procedure. Although
exclusion of wetland plants has the disadvantage of lowering
the pollen sums, this procedure yielded five statistically sig-
nificant LPAZs (Figs.5, 6). In PT2 the same analysis, with
wetland plants included, revealed two statistically significant
LPAZs (Fig.7). Microscopic charcoal particles (> 10µm)
were analysed in core PT1 following standard methods and
presented as concentrations and influxes (Fig.6; Tinner and
Hu 2003; Finsinger and Tinner 2005). All palynological data
derived from this study are presently stored in the Alpine
Palynological Database (ALPADABA).
The programs CANOCO 4.5 and CanoDraw 4.14 (Lepš
and Šmilauer 2003) were used for ordinations. For the
vegetation survey and the surface samples, detrended cor-
respondence analysis (DCA) was carried out with detrend-
ing by second order polynomials (Fig.8). Since the gradi-
ent lengths of DCA axis 1 were smaller than 2.5 standard
deviations in PT1 and PT2 (Legendre and Birks 2012),
we decided to use the linear response model of principal
component analysis (PCA) for the records (Figs.5, 8, 9).
Fig. 4 Age-depth diagram for “Palm Terrace” 1 (PT1). Black dots
represent the three calibrated radiocarbon ages from macroscopic
charcoal (Table1), with error bars for vertical thickness of the sam-
ples and 2σ for estimated age (Calib 7.1, Reimer etal. 2013). The
model (smooth spline 0.3, black line) was developed with the pro-
gram clam 2.2 (Blaauw 2010), the outer lines show the 95% confi-
dence envelope of the generalized mixed-effect regression (Birks and
Heegaard 2003). The red dot shows the dated rhizome that was not
included in the model calculations

Vegetation History and Archaeobotany
1 3
X‑ray fluorescence (XRF) analysis, loss onignition
andconductivity
XRF analysis was performed on PT1 and PT2 at the Institute
of Geological Sciences, University of Bern, using an Itrax
core scanner (Cox Analytical System) equipped with a Mo
tube (30kV, 40mA) at 1cm resolution and an analysis
time of 20s per measurement. The lines were smoothed by
applying a 5 point moving average, and only values with a
mean squared error (MSE) lower than 3 were taken into con-
sideration. The top 50cm sediment of PT1 was accidentally
lost before analysis, so that XRF and other data are missing
for this part.
Loss on ignition (LOI) was performed on 23 samples
of 1cm3 from PT1 and 8 from PT2, after drying the sam-
ples > 24h at 105°C, in a muffle furnace at 550°C and
then 950°C (Heiri etal. 2001). To estimate changes in total
dissolved solids, relative conductivity measurements were
performed on 2cm3 sub-samples from 35 surface samples,
seven samples from PT1 and eight samples from PT2. These
were diluted in 20–400ml of distilled water and the con-
ductivity of the solution was then measured (HANNA HI
98129) and converted to total dissolved solids (TDS) in the
solution, as parts per million (ppm) per 1cm−3 of sediment
(Figs.3, 9).
Results andinterpretation
Vegetation survey andsurface samples
The map of the oasis (Fig.3) together with the ordination of
the vegetation survey (Fig.8a) show spatial and statistical
separation into two main vegetation belts: wetland vegeta-
tion in the mire with T. domingensis and P. australis, and
drought-adapted upland vegetation around the mire with
Nitraria retusa and Chenopodiaceae. In between these two
vegetation types were scattered P. dactylifera palms. In
large oases elsewhere, Phoenix can form a mesophilous date
palm belt between wetland and drought-adapted vegetation
(Horowitz 1992). The moisture gradient at “Palm Terrace”
is probably too steep and spatially too undefined to allow a
distinct palm belt to grow, but the spatial distribution of the
single palms between the wetland and the drought-adapted
vegetation (Fig.3, green triangles) argues for the presence of
2000
1950
1900
1850
1800
1750
1700
1650
1600
1550
1500
1450
1400
1350
1300
1250
1200
1150
1100
Age [cal A.D.]
10
20
30
40
50
60
70
80
90
100
110
120
140
160
180
Depth [cm]
230±50
690±40
770±70
Rad
iocarbon dates [yr uncal B.P.]
20
Phoenix dactylifera
Pinus
Salix
Fagus
Quercus
Castanea sativa
Olea
Casuarina
Tamarix
Nitraria
cf. Capparis
20 40 60 80 100
Trees
Shrubs
Herbs
Wetland plants
Achillea-type
Artemisia
Aster-type
Brassicaceae
Caryophyllaceae
Centaurea
Centaurium
Cerealia-type
20 40 60
Chenopodiaceae
Cichorioideae
Helianthemum
Lythrum
Mentha-type
Paronychia
Plantagocoronopus-type
Plantago
Plantago lanceolata-type
Plantago maritima-type
20 40 60
Poaceae
Salsola-type
Trifolium-type
20 40 60
Typha domingensis-type
Cladium
Cyperaceae
Alisma-type
Pseudoschizaea
50 100
Lasiodiplodia
Glomus
Microcharcoal conc. [frag./cm3]
Microcharcoal influx [frag./cm2/yr]
400
Pollen sum
LPAZ
PT1-4
PT1-3
PT1-2
PT1-1
PT1-5
ignuFeaglAstnalpdnalteWsbreHsburhSseerT
Lithology
Decomposed Peat Rhizome Peaty Gyttja Silty Clay GyttjaSilty Clay
20 6040
PCA 2nd axis
PCA 1st axis
20 60 100
x10'000x1'000
100 200024 0240101010101010101010101010101010101010101010101010101010101010101 10 10
% terrestrial pollen sum
2000 4000
Total pollen influx [grains/cm2/yr]
Spirogyra
Fig. 5 Lithology, pollen and spore percentages, microscopic charcoal
concentrations (particles > 10µm/cm3) and influx (particles/cm2/year)
from “Palm Terrace” 1 (PT1). Sample scores of PCA axis 1 and 2 are
used to summarize the vegetation dynamics. Empty curves show ×10
exaggerations. LPAZ Local pollen assemblage zone (statistically sig-
nificant when excluding wetland plants)
▸

Vegetation History and Archaeobotany
1 3
a perhaps relict P. dactylifera ecotone zone (see vegetation
history section).
The mire vegetation (Fig.3, magenta/violet circles, ESM
Fig.1) was dominated by Saccharum ravennae, P. austra-
lis and in the wettest places T. domingensis, all of which
are hydrophytes (Danin and Orshan 1999). These species
are common indicators of flowing water (Danin and Orshan
1999). S. ravennae and T. domingensis are characteristic
of fresh water, whereas P. australis also grows in slightly
saline water (Danin 1983). Outside the mire under meso-
philous conditions, P. dactylifera, Tamarix nilotica, Juncus
maritimus and Imperata cylindrica were the most common
species. Nitraria retusa formed an eponymous belt at the
outermost limit of the oasis vegetation (Fig.3, blue squares).
It is drought-adapted and highly salt-tolerant (up to 56 dS/m,
Al-Oudat and Qadir 2011), as is also the more mesophil-
ous T. nilotica, which can accumulate salt in its leaves and
excrete it again, which can then increase the salinity of the
surface soil (Decker 1961; Mozingo 1987; Al-Oudat and
Qadir 2011). Drought-adapted Chenopodiaceae are the
dominant plants in the upland vegetation around the mire.
Desert springs are considered to be the natural habitat
of the mesophilous date palm, P. dactylifera (Danin 1983).
The palm is reported to need a high freshwater table (Danin
and Orshan 1999), although it frequently grows on moist
saline soil with a salt crust, if the roots can penetrate to less
saline water beneath (Danin 1983). The distribution of the
vegetation together with the topography of the oasis points
Fig. 6 Comparison between pollen percentages (black), concentrations (pollen grains/cm3, dark grey) and influx (pollen grains/cm2/year, light
grey) of selected pollen types from core PT1. Empty curves show ×10 exaggerations
Fig. 7 Lithology, selected
pollen and spore percentage dia-
gram from core “Palm Terrace”
2 (PT2) plotted on a depth scale.
Empty curves show ×10 exag-
gerations. LPAZ local pollen
assemblage zone (statistically
significant including wetland
plants)

Vegetation History and Archaeobotany
1 3
to an inflow of fresh water close to the eastern border of the
mire. This is confirmed by finding a spring with many young
seedlings of P. dactylifera close to the eastern limit of the
mire (Fig.3). In general, a negative relationship between
percentage seed germination and salt concentration has been
observed for P. dactylifera (Ramoliya and Pandey 2003;
Alhammadi and Kurup 2012).
Date palms are dioecious, and only a few pollen-pro-
ducing male trees are needed for date production, i.e. a
male–female ratio ~ 1:50 (Jain etal. 2011; Chao and Krue-
ger 2007). Uncultivated stands are characterized by a higher
proportion of males, the occurrence of trees with cluster-
ing trunks (Jain etal. 2011) and abundant J. maritimus var.
arabicus, which is also very salt tolerant (> 56 dS/m, Al-
Oudat and Qadir 2011), in the undergrowth (Danin 1983).
The palm stands in the marginal zone of “Palm Terrace”
fulfilled these criteria to a fair degree, suggesting that these
dates were not cultivated for food.
DCA axis 1 of the vegetation survey shows a water avail-
ability gradient and groups the wetland plants T. domingen-
sis, Eupatorium cannabinum, Lythrum salicaria, S. raven-
nae and P. australis opposite to N. retusa, Chenopodiaceae
and “open ground”, whereas mesophilous taxa such as T.
nilotica, P. dactylifera and J. maritimus are located in an
intermediate position (Fig.8a). On DCA axis 2 T. nilotica
is distant from the rest, pointing to a salinity gradient within
the mesophilous vegetation. Conductivity measurements on
the surface soil samples indicate an increasing salinity from
the centre towards the edge of the oasis with increasing
salinity down the slope and with increasing distance from
the freshwater inflow at its eastern edge (Fig.3).
Modern pollen surface samples together with present-
day vegetation surveys are the key for the interpretation
of past vegetation changes from pollen sequences (Baruch
1993; Davies and Fall 2001; Lopez-Merino etal. 2016).
The results from six pollen surface samples are presented as
pollen percentages (Fig.3) as well as concentrations (ESM
Fig.2) and compared to the present-day vegetation using
Fig. 8 Species plots from the ordination analysis. a DCA of the veg-
etation surveys, axis 1 explains 22.8% of the variance in the data set
and axis 2 explains 13.7%. “Chenopodiaceae + indet” includes Che-
nopodiaceae/Amaranthaceae and other plants that were unidentifia-
ble, flowerless and often desiccated. (P.) after the name stands for the
family of Poaceae, (C.) for the sub-family Cichorioideae. b DCA of
the surface pollen samples, where axis 1 and 2 explain 59 and 11.9%
of the data variance, respectively; the star represents a group of pol-
len types that only appear in one sample, namely Allium, Centaurea,
Centaurea scabiosa-type, Cucumis sativa, Daucus, cf. Frankenia,
Plantago maritima-type, Polygonum alpinum-type, Quercus, Salix
and Solanum nigrum. c PCA of the pollen samples from PT1, 1st axis
explains 49% of the variation, 2nd axis 28.3%, only the pollen types
with values > 0.2 are shown, as well as the microscopic charcoal
influx; the following group of pollen types are represented as a spiral:
Alisma-type, Astragalus, Calendula, Carlina and Cerealia-type
▸

Vegetation History and Archaeobotany
1 3
ordination (Fig.8a, b). On DCA axis 1, Chenopodiaceae/
Amaranthaceae (Cheno/Ams) pollen figures as the main
opponent to T. domingensis-type and Poaceae (including P.
australis and S. ravennae) pollen (Fig.8b). In samples b,
d and e, dominance of Cheno/Ams pollen (up to 98% in b,
Fig.3) may have suppressed the percentage representation
of other taxa, considering that concentrations of these taxa
are relatively high (ESM Fig.2). 27.5% of all halophytic
taxa in the Mediterranean basin belong to the Chenopo-
diaceae family (Shaer and Squires 2015). The interpreta-
tion of this family is, however, problematic due to the huge
variety of different taxa which grow in the region, includ-
ing annual herbs but also shrubs with distinct ecologies
(Zohary and Feinbrun-Dothan 1966; Horowitz 1992). In
the pollen surface samples, the highest concentration of T.
domingensis-type pollen was found in the innermost sample
of the mire (d), where neither N. retusa nor T. nilotica pol-
len were detected (Fig.3). The most abundant pollen types
in the outermost samples on the northern side of the oasis
were Cheno/Ams, N. retusa, T. nilotica and P. dactylifera
(Fig.3a, b). These results show that pollen assemblages from
oasis surface samples generally represent local vegetation
accurately. If dated, sediment samples from deeper layers
may thus provide useful insights into long-term vegetation
dynamics. It is important, however, to keep in mind that
Cheno/Ams might be somewhat overrepresented in the local
pollen records due to the great abundance of the plants in
the surrounding upland vegetation. Conversely, certain taxa
such as Nitraria and Tamarix might be underrepresented.
Chronology
The age-depth model for core PT1 (Fig.4) is based on three
calibrated 14C dates from charcoal samples (Table1) and
dates back to ca. ad 1100, although the 95% confidence
envelope, which considers both age and depth uncertainties
(Birks and Heegaard 2003), indicates that the bottom of the
record might be significantly older (ad 390). Our interpreta-
tions are based on the most likely age (Fig.4, black line),
but they consider the large uncertainties that are inherent in
the dating approach. During the past ~ 1,000years the model
suggests that there were only small changes in sedimenta-
tion rates. From bottom to top, sedimentation rates are rela-
tively high (0.4cm year−1) until 140cm depth, when they
decrease to 0.1cm year−1 at 80cm depth and then slightly
increase again to 0.2cm year−1 at the top of the core. This
is also reflected in pollen concentrations, with low values in
the lowermost part of the record (Fig.6). No date could be
obtained from core PT2 due to the lack of suitable terrestrial
macrofossils.
“Palm Terrace” 1, vegetation andfire history
according topollen andcharcoal
The pollen diagram of core PT1 shows five local pollen
assemblage zones (LPAZ, Figs.5, 6, 9; ESM Fig.3). The
pollen zones are statistically significant when pollen from
wetland plants is excluded from the pollen sum. This find-
ing suggests that only minor changes occurred in wetland
Fig. 9 X-ray fluorescence (XRF) raw counts; titanium, potassium,
zirconium, silicon, silicon/titanium ratio, calcium, chlorine and bro-
mine, relative conductivity (TDS), loss on ignition (LOI) and Princi-
pal Component Analysis (PCA) from the pollen data of the two cores
PT1 (red) and PT2 (blue) plotted against depth. Dashed lines indicate
local pollen assemblage zones (LPAZ) from PT1

Vegetation History and Archaeobotany
1 3
vegetation, whereas upland vegetation significantly varied
over time. Poaceae are not considered as limited to wetlands
because some species such as I. cylindrica were also com-
mon outside the mire. Nevertheless, we assume that most
Poaceae pollen derives from P. australis and S. ravennae,
both of which were dominant in the mire.
In LPAZ PT1-1 (ad ~ 1100 to ~ 1130), high percentages
of herb pollen, mainly Cheno/Ams and Cichorioideae, point
to a well-developed upland vegetation belt. Phoenix dac-
tylifera pollen indicates the presence of mesophilous date
palm stands. The high percentage of T. domingensis-type
(up to 50%) suggests that it dominated the wettest part of the
mire. The extremely high values of Glomus fungal fruiting
bodies may derive from soil erosion (Anderson etal. 1984;
van; Geel 1986; Kołaczek etal. 2013). Low microscopic
charcoal concentrations and influx values suggest low fire
activity in the region.
The continuous and increasing pollen curve of T. nilotica
as well as the first grains of N. retusa suggest an expan-
sion of salt-tolerant vegetation with shrubs in LPAZ PT1-2
(ad ~ 1130 to ~ 1220). Single finds of Capparis pollen indi-
cate the presence of this typical Mediterranean shrub. Per-
centages of over 10% Glomus fruiting bodies again point to
soil instability. Pollen percentages of P. dactylifera remain
at ca. 10%, suggesting unchanged abundance of this palm in
the oasis. Wetland plant sums are lowest in this zone, indi-
cating lower water availability compared to the rest of the
sequence. Lasiodiplodia fungal spores are abundant, point-
ing to infestation by this plant pathogen at or close to the
site. The slight rise in microscopic charcoal suggests that fire
activity marginally increased in the catchment of the oasis.
LPAZ PT1-3 (ad ~ 1220 to ~ 1320) is characterized by
a strong rise in Cheno/Ams pollen percentages, peaking in
this pollen zone. This suggests expansion of drought-adapted
upland vegetation, mainly at the cost of wetland vegetation,
which has low values, and mesophilous T. nilotica (Fig.6).
In agreement, pollen data suggest that the drought-adapted
N. retusa expanded markedly, while mesophilous P. dac-
tylifera contracted to minimum abundances (percentages,
concentrations and influx). The rise in microscopic charcoal
suggests increasing fire incidence.
In PT1-4 (ad ~ 1320 to ~ 1650) T. domingensis-type and
Poaceae markedly increase, indicating a distinct expan-
sion of wetland vegetation. Towards the end of this zone at
around ad 1600, percentages of wetland plants rise further,
reaching their highest values around ad 1700, pointing to
a marked increase of moisture availability throughout this
period. This interpretation is reinforced by a progressive
expansion of P. dactylifera and a decline of Chenopodiaceae,
suggesting that the moisture-requiring date palms expanded
into the drought-tolerant vegetation belt of the oasis.
Pollen zone PT1-5 (ad ~ 1650 to today) is characterized
by a major increase of moisture-demanding P. dactylifera
percentages, suggesting a strong expansion of date palms
in the oasis that peaked around ad 1850. This is also very
well seen in both concentration and influx values, showing
that this major vegetation change was real and not related
to distortions from percentage calculation. The dominance
of date palm vegetation, probably forming its own belt at
“Palm Terrace”, was preceded by the expansion of wetland
vegetation around ad 1700–1750. Around ad 1750–1800,
when conditions were moist, arable farming may have
been practised at the oasis, as shown by single Cerealia-
type and Plantago maritima-type finds. Towards the end
of the zone, at ad 1900–1950 date palms declined, while
fire activity and soil erosion markedly increased as shown
by a microscopic charcoal peak and an increase in Glomus
fungal fruiting bodies (Fig.5).
Taken together, the pollen record of PT-1 shows an
increasing expansion of moisture-requiring vegetation
such as T. domingensis-type and P. dactylifera that came
to an end during the twentieth century. These dynamics
are summarized by the ordinations. In the taxa scores, the
PCA axis 1 of PT1 (explaining 49% of the variance) spans
along a gradient from drought-adapted Chenopodiaceae to
moisture-requiring P. dactylifera, T. domingensis-type and
Poaceae (probably dominated by P. australis and S. raven-
nae), and may thus primarily reflect a moisture gradient
(Fig.8c). Scores of PCA axis 1 through time can therefore
be used as a moisture indicator curve (Fig.5). PCA axis 2
(28%) mainly spans between Poaceae and T. domingensis-
type, which is more difficult to explain. However, aligned
along depth and thus with time (Fig.5) the sample scores
of PT1 show that PCA axis 1 is mainly related to Chenopo-
diaceae, while PCA axis 2 closely follows T. domingensis-
type. The pollen records of these two plant taxa can thus
be used to illustrate the primary vegetation dynamics of
the oasis.
“Palm Terrace” 2, vegetation history frompollen
Compared to PT1, PT2 is fragmentary and undated (Fig.7).
Nevertheless, it can be used to assess whether major vegeta-
tion patterns found at PT1 can also be detected at the edge
of the oasis. Indeed, PT1 and PT2 share common biostrati-
graphic trends such as high Poaceae pollen percentages that
decrease towards the top and a steady increase of P. dactyli-
fera pollen from 160 to 45cm (LPAZ PT2-1). The drastic
increase of Cheno/Ams pollen (up to 98%) in the topmost
45cm (LPAZ PT2-2) is associated with an almost complete
disappearance of date-palm vegetation. Although more pro-
nounced, this marked change may correspond to the date
palm decline of the twentieth century in PT1, which was
also linked to a minor subsequent increase of Cheno/Ams
(Figs.5, 6, 7).

Vegetation History and Archaeobotany
1 3
Lithology, XRF, conductivity andloss onignition
asenvironmental proxies
The lowermost part of core PT1 consisted mainly of silty
clay, changing to a mixture of silty clay gyttja with increas-
ing organic matter towards the top, where it gradually
changed from peaty gyttja to decomposed peat (Fig.5). The
bottom of PT2 consisted mainly of sandy silt (246–230cm)
and silty clay (230–225cm), which changed progressively
into silty clay gyttja (190–165cm; Fig.7). Towards the
upper part of the core, prominent changes in sedimentary
composition could be observed, starting with peaty gyttja,
overlain by silty clay gyttja up to a sand layer at the top
that in some parts contained peaty constituents (Fig.7). The
greater number of lithological changes in PT2 suggests that
the edge of the oasis may have been more sensitive to envi-
ronmental fluctuations. In both cores, thick rhizomes were
found; in PT1 at 60–64cm (dated to recent times, Fig.4;
Table1) and in PT2 at 67.5–71cm depth.
The results of XRF measurements (Fig.9) show that
lithogenic elements such as titanium, potassium and zirco-
nium generally decrease in values from bottom to top in both
cores. This pattern is probably caused by a dilution of detri-
tal components by increased organic matter contents, per-
haps associated with the spread of palm tree vegetation. Sili-
con shows the same process, with one major difference at the
top of PT2 where values rapidly increase from ~ 33cm depth
upwards (data for the upper 50cm are missing for PT1).
The silicon/titanium ratio (Fig.9) shows that the diverging
of the two curves in PT2 started already at ~ 70cm, while
the ratio remains stable for PT1 at least up to 50cm depth.
Diverging of silicon from other detrital elements may be
explained by a second source for silicon, possibly originat-
ing in the surrounding sandstones and brought to the oasis,
either by wind or by surface runoff. The sand layer in PT2
supports this hypothesis. The calcium curve of PT1 shows
only minor fluctuations, although that of PT2 shows higher
values at both the bottom and top. Calcium is difficult to
interpret, since it can occur in many forms such as gypsum
(CaSO4·2H2O) and calcium carbonates (CaCO3) (Vepraskas
and Craft 2016). Calcium accumulations, however, may be
linked in appropriate contexts to increased desiccation and/
or lower water flow causing water supersaturation. Chlo-
rine and bromine counts generally increase towards the top
over the whole sequences of both cores. These are salinity
indicators, and an increase can result from either higher salt
content in the spring water, enhanced evaporation, or a com-
bination of both (Croudace and Rothwell 2015; Beffa etal.
2016). The conductivity curves of PT1 and PT2 (Fig.9)
show the same increasing trend towards the top regarding
chlorine, with higher values in PT2, in our relative measure-
ments, up to 1,158ppm per cm3 of sediment. Compared to
the conductivity of the surface samples (Fig.3), especially
those from the drought- and salinity-adapted Nitraria belt
reaching up to 23,355ppm per cm3 of sediment, the values
of the cores are still at a quite low level, and indicate increas-
ing salt content towards both the soil surface and the edges
of the oasis. Organic matter (OM) measured through loss on
ignition (LOI 550°C, Fig.9) of PT1 increases towards the
top up to 18.7%, while in PT2 it shows large oscillations.
LOI at 950°C from PT1 shows constant amounts of car-
bonates or potentially gypsum at low level, while in PT2 a
sharp increase is shown at ~ 60cm (there is a corresponding
rise of calcium in the XRF measurements) suggesting more
evaporation (Tiner 1999), followed by a decrease.
Multi‑proxy summary interpretation
Both cores PT1 and PT2 show an increase of arboreal pol-
len, mainly from palms in the last millennium, before a
sharp reduction in the last ca. 100years (Figs.5, 7, 10).
Conversely, detrital elements such as titanium and silicon
continuously decrease in the last millennium, before silicon
drastically increases in the last century (Fig.9), suggesting
Fig. 10 Comparison of regional palaeoclimatic indicators with veg-
etation, erosion and fire dynamics at “Palm Terrace”. a δ18O values
from Soreq Cave speleothems as an indicator of past rainfall in the
eastern Mediterranean (Bar-Matthews et al. 2003); b water level
reconstructions of the Dead Sea from Bookman et al. (2004) (red
line) and Migowski etal. (2006) (orange line); c pollen percentages
of Phoenix dactylifera (date palm) as an indicator of wetland expan-
sion; d percentage values of Glomus fungal fruit bodies as an indi-
cator for soil erosion; e microscopic charcoal influx representing
regional fire activity

Vegetation History and Archaeobotany
1 3
that plant growth may have controlled the amount of min-
eral material brought in by erosion. Given that the presence
of trees and shrubs in an oasis is primarily determined by
moisture availability, these coupled bio-geosphere dynamics
ultimately reflect changes in the water table. The significant
change at ~ 70–60cm depth in PT2 (ca. ad 1600–1700),
at the edge of the mire, when silicon and titanium records
diverge, with greater amounts of silicon, indicates an addi-
tional source of detrital input to the wetland, as well as a
potential increase in evaporation suggested by higher cal-
cium contents (PT2, Fig.9). Chlorine counts and relative
conductivity values rise drastically in the last 100 years,
tracking pollen PCA axis 1 (Figs.8c, 9), which is mainly
controlled by the drought- and salinity-adapted Chenopodia-
ceae (Figs.5, 6, 8, 9). At the same time, the mesophilous
palm vegetation disappears, indicating a drastic change in
vegetation. If we assume similar ages at similar depths in
PT1 and PT2, as confirmed by the generally good litho- and
chemostratigraphical (XRF) agreement between the two
sequences, the wetter centre of the mire (core PT1) seems
to have reacted somewhat later and less drastically, with a
moderate Chenopodiaceae expansion and a less pronounced
reduction of palm (Figs.5, 6) than the drier edge in core
PT2. The persistently higher abundances of the moisture-
demanding halophyte Tamarix in core PT2 (Fig.7) match
the higher salinity reconstructions for this site at the drier
edge of the oasis, as inferred from the high relative conduc-
tivity values and greater chlorine abundances (Fig.9).
In summary, the multi-proxy evidence from PT1 and
PT2 points to a drastic shrinkage of the mire and the oasis
in the last ca. 100 years, probably due to less water inflow
from the spring and possibly to greater evaporation shown
by increased carbonate and salt contents; this is potentially
linked to increased dust and/or input of eroded mineral mate-
rial indicated by additional silicon-dominated detrital input,
and also corroborated by the increased occurrence of Glo-
mus fruiting bodies. Whereas the moisture availability was
still high enough to support the mire and to some extent
the palm vegetation in or close to the centre of the wetland
(PT1), the mire plants and palms at the edge (PT2) were
almost completely replaced by halophytes such as Cheno-
podiaceae, N. retusa and J. maritimus.
Discussion
The palynological and lithological evidence suggests that
the oasis vegetation was confined to a rather small area
before ad ~ 1300 (Figs.5, 6, 9, LPAZs PT1-1, PT1-2, PT1-
3), to subsequently expand until ca. ad 1400–1500. Oasis
vegetation temporarily contracted around ad 1500–1600,
to re-expand markedly around ad 1700–1900 before the
final contraction, which occurred during the past century.
These oasis dynamics were most probably driven by local
moisture availability.
Reconstructed water levels of the Dead Sea may be
used as a proxy for regional moisture dynamics, given
that these were primarily responding to changes in the
balance between evaporation and precipitation (Enzel etal.
2003; Dayan and Morin 2006; Robinson etal. 2006; Bar-
tov etal. 2007; Rambeau and Black 2011). High-resolution
reconstructions by Bookman etal. (2004) and Migowski
etal. (2006) (Fig.10) suggest that Dead Sea water levels
increased at ad ~ 1200, with a high level at ~ 1350, then
they declined between ~ 1400 and 1600, increased again
at ~ 1900 and markedly declined to 418m b.s.l. by 2005,
and then to 430m b.s.l. in 2015 (Fig.10; Yechieli etal.
2016). In general, these reconstructions agree well with
δ18O based reconstructions of rainfall from the Soreq cave
speleothems, mineral deposits formed in caves (Fig.10;
Bar-Matthews etal. 2003). Considering the chronologi-
cal uncertainties of our records, it is most likely that
these regional moisture changes have a direct link to the
environmental dynamics reconstructed at “Palm Ter-
race”. In particular, increased regional moisture avail-
ability would explain the marked expansions of the oasis
at ad ~ 1300–1500 and ~ 1700–1900 and the decreased
regional moisture availability around ~ 1400–1600
would explain the contractions of the oasis wetland at
~ 1500–1700 (Figs.5, 6, 10).
Available vegetation history reconstructions for the
region (Baruch 1993; Heim etal. 1997; Neumann etal.
2007, 2009, 2010; Leroy 2010; Leroy etal. 2010; Litt etal.
2012; Langgut etal. 2014, 2015) come from the Dead Sea
itself, not from local oases. The surface of the Dead Sea is,
however, so large (620km2 in 2012; Ghatasheh etal. 2013)
that it reflects the vegetation dynamics of a much larger
regional area (Moore etal. 1991; Lang 1994), including the
Mediterranean biome. Another difference is that the Dead
Sea receives part of its pollen from inflowing rivers (catch-
ment: 40,650km2, Klein and Flohn 1987), so that a regional
signal with prominent Q. calliprinos, Pistacia and Olea is
common (Leroy 2010; Neumann etal. 2010). The sediment
record from “Palm Terrace” oasis, on the other hand, con-
tains, with rare exceptions, only pollen of local origin, and
has therefore a markedly different vegetational history, due
to the absence of Mediterranean trees in the oasis. However,
climatic effects on the vegetation in the Dead Sea area were
reconstructed by Neumann etal. (2007) by investigating two
sediment records from Ein Feshka and Ze’elim (Fig.1). Rel-
atively moist conditions favouring agriculture were inferred
for the period ca. ad 1200–1500, followed by a short dry
period after 1500, when these records end. These results also
generally agree well with the phases of oasis expansion and
wetland contraction at “Palm Terrace” at ~ 1300–1500 and
~ 1500–1700, respectively.

Vegetation History and Archaeobotany
1 3
The final contraction of the oasis and the expansion of
drought- and salinity-adapted vegetation (predominantly
Chenopodiaceae) towards the top of our records in the centre
(PT1) and particularly at the edge (PT2) of the mire might
be closely connected to the drop of Dead Sea water levels
which started during the twentieth century (Bookman etal.
2004; Migowski etal. 2006; Bartov etal. 2007). Lowering of
Dead Sea water levels during recent decades was caused by
abstraction of water from the river Jordan and from aquifer
reservoirs mainly for irrigation, and from the Dead Sea itself
for industrial mineral extraction (Waitzbauer 2004; Lensky
etal. 2005; Abu Ghazleh etal. 2009, 2011; Chen and Weis-
brod 2016; Siebert etal. 2016). The shrinkage of the Dead
Sea probably causes a positive feedback of aridification,
since the smaller water surface results in less evaporation
from it, which reduces the air humidity in the surrounding
areas and thus increasing the aridity in the region (Salameh
and El-Naser 2009; Flexer and Yellin-Dror 2009; Abu Gha-
zleh etal. 2011). On the basis of precipitation measurements
in the region from 1970 to 2002, Kafle and Bruins (2009)
suggested slightly increasing trends of precipitation at the
Mediterranean coast of Israel, but decreasing trends further
inland. Specifically, a statistically significant decrease was
observed at the meteorological station of Sedom Pans, lying
next to the southern Dead Sea basin. In agreement, all Jor-
danian meteorological stations show decreasing precipita-
tion trends during the last decades (Freiwan and Kadioğlu
2007). Furthermore, an abrupt and significant increase in
both the mean, minimum and maximum temperatures have
been observed since 1957 and 1967 (Smadi 2006), as well
as a decrease in the diurnal temperature range (Freiwan and
Kadioğlu 2007).
No archaeological evidence of human occupation is
known that would help to assess the human legacy at the
site. The map of archaeological sites around the Dead Sea
provided by Beit-Arieh (1997) shows a relatively densely
populated western shore, whereas on the eastern shore only
a few sites at some distance north and south of the oasis
are mapped. The closest archaeological finds come from
Kallirrhoë, ~ 7km north of “Palm Terrace” (Fig.2), which
was used as a thermal bath during Roman times, known
during Byzantine times and rediscovered by Ulrich Jaspar
Seetzen in ad 1807 (Seetzen 1854; Strobel 1977; Clamer and
Dussart 1997; Clamer 2010). Some of the rivers described
by Seetzen are, according to Kruse (in Seetzen 1854), the
“Bäche Pisga” and one of them probably corresponds to
the, at that time more active, “Palm Terrace” oasis. Detailed
population distribution data from 1596 shows Mazra’a as
the closest settlement (14 households, ~ 70 persons) about
20km south of Wadi Mujib and Ras (13 households, ~ 65
persons) further to the east, but neither nomadic tribes nor
settlements in the area of the oasis are known (Hütteroth and
Abdulfattah 1977). The lack of archaeological evidence, a
drastically reduced population in the area, due to the cru-
saders and the Black Death plague (Dols 1977; Broshi and
Finkelstein 1992; Hillenbrand 2000) together with the harsh
local environment let us speculate that the “Palm Terrace”
area was not intensively used by humans until ~ 1800, when
pollen of crops (Cerealia-type), ruderals (Plantago) and a
marked increase of fire activity (charcoal) in PT1 point to
human activities (Fig.5). Of special interest is the pollen
of Casuarina, which was found in the uppermost level; this
drought-adapted tree was imported from Australia for eco-
nomic purposes and is also reported in other pollen studies
of the area (Horowitz 1979; Baruch 1986; Heim etal. 1997;
Leroy 2010; Litt etal. 2012; Schiebel 2013).
During the past century, the human population of Jordan
has increased enormously, from 225,330 inhabitants in 1922
to 7,748,000 in 2016 (Barham 1994; UN data 2016), and
more water such as that of the rivers Jordan and Yarmouk
was extracted, dammed or channelled; such works include
the Degania dam in 1932, the King Abdullah canal in 1964
and the Unity dam in 2011 (Borchardt etal. 2016). However,
we cannot assess whether the shrinkage of the “Palm Ter-
race” oasis was caused primarily by reduced precipitation
(Kafle and Bruins 2009), or if the main driver was the Dead
Sea lowering that in turn resulted in a lower ground water
table in the area (Salameh and El-Naser 2000; Abu Gha-
zleh etal. 2011). Such a change could have caused less flow
from the oasis springs (Weinberger etal. 2012; Borchardt
etal. 2016) and ultimately the contraction of its vegetation.
It is thus likely that land use contributed to exacerbate the
effects of decreasing precipitation during the past 100 years,
drastically changing the extent of the “Palm Terrace” oasis.
Our results unambiguously demonstrate the high climatic
sensitivity of oasis environments. Future climate in Jordan
and elsewhere in the eastern Mediterranean region will prob-
ably feature a further temperature increase and a decrease
in precipitation (Black etal. 2011; Christensen etal. 2013;
Lelieveld etal. 2016). Together with continued water extrac-
tion, this may severely endanger oasis environments such
as that of “Palm Terrace”, possibly leading to their final
disappearance.
Conclusions
Surface pollen studies of oases can significantly contribute
to refine vegetation history reconstructions, when linked to
modern vegetation surveys. The oasis vegetation at “Palm
Terrace” underwent several phases of expansion and con-
traction, which can be connected to previously reconstructed
moisture changes in the Dead Sea area. The combination
of exponentially increasing water extraction and decreased
precipitation has strongly affected the environments of the
oasis since the beginning of the twentieth century, leading

Vegetation History and Archaeobotany
1 3
to a drastic contraction of the wetland. So far it is unclear
whether other oases in the area suffered similar develop-
ments. Multi-proxy investigations at other oasis sites might
provide further evidence of a connection between local con-
ditions and regional trends. Our study shows that oasis stud-
ies have a high potential for clarifying vegetation dynamics
at the local scale. For instance, rates of change or time lags
of different taxa such as P. dactylifera and T. domingensis
might be explored in detail, generating high-resolution and
-precision multi-proxy series. Taxonomic improvements
may help to distinguish important taxa within the Chenopo-
diaceae/Amaranthaceae family. Such improvements, as well
as the study of other wetlands in arid contexts, may gener-
ate important knowledge to anticipate future oasis dynamics
under conditions of global change.
Acknowledgements We gratefully thank the Jordan Royal Society
for the Conservation of Nature for the permission to do fieldwork in
the Mujib Biosphere Reserve. Furthermore, we are grateful to Steffen
Wolters for his help with the fieldwork, Sönke Szidat for radiocarbon
dating and Jean-Nicolas Haas and Walter Gams for their help with
identification of the fungal spores. We would also like to thank two
anonymous reviewers and the handling editor for constructive com-
ments that greatly improved this manuscript. Funding was provided by
the Swiss National Science Foundation (Grant-Nr. 136731).
References
Abu Ghazleh S, Abed AM, Kempe S (2011) The dramatic drop of the
Dead Sea: background, rates, impacts and solutions. In: Badescu
V, Cathcart RB (eds) Macro-engineering seawater in unique envi-
ronments. Springer, Berlin, pp77–105
Abu Ghazleh S, Hartmann J, Jansen N, Kempe S (2009) Water
input requirements of the rapidly shrinking Dead Sea.
Naturwissenschaften 96:637–643. https://doi.org/10.1007/
s00114-009-0514-0
Albert R, Petutschnig B, Watzka M (2004) Zur Vegetation und Flora
Jordaniens. In: Enzel Y, Agnon A, Stein M (eds) New frontiers
in Dead Sea paleoenvironmental research. Geological Society of
America, Boulder, pp215–229
Alhammadi MS, Kurup SS (2012) Impact of salinity stress on date
palm (Phoenix dactylifera L.)—a review. In: Sharma P, Abrol V
(eds) Crop production technologies. InTech, Winchester, pp169–
178. https://doi.org/10.5772/29527
Al-Oudat M, Qadir M (2011) The halophytic flora of Syria. Interna-
tional Center for Agricultural Research in Dry Areas, Aleppo
Anderson RS, Homola RL, Davis RB, Jacobson GL Jr (1984) Fossil
remains of the mycorrhizal fungal Glomus fasciculatum complex
in postglacial lake sediments from Maine. Can J Bot 62(2):325–
322,328. https://doi.org/10.1139/b84-316
Barham N (1994) Entwicklung und Probleme der Landwirtschaft Jor-
daniens. Zeitschrift für Wirtschaftsgeographie 38:23–35
Bar-Matthews M, Ayalon A, Gilmour M, Matthews A, Hawkesworth
CJ (2003) Sea-land oxygen isotopic relationships from plank-
tonic foraminifera and speleothems in the Eastern Mediterranean
region and their implication for paleorainfall during interglacial
intervals. Geochim Cosmochim Acta 67:3,181–3,199. https://doi.
org/10.1016/S0016-7037(02)01031-1
Bartov Y, Enzel Y, Porat N, Stein M (2007) Evolution of the Late Pleis-
tocene Holocene Dead Sea basin from sequence stratigraphy of
fan deltas and lake-level reconstruction. J Sediment Res 77:680–
692. https://doi.org/10.2110/jsr.2007.070
Baruch U (1986) The Late Holocene vegetational history of Lake Kin-
neret (Sea of Galilee). Israel Paléorient 12:37–48. https://doi.
org/10.3406/paleo.1986.4407
Baruch U (1993) The palynology of Late Quaternary sediments of
the Dead Sea. The Hebrew University of Jerusalem, Hebrew (in
Hebrew with English abstract)
Beffa G, Pedrotta T, Colombaroli D etal (2016) Vegetation and fire
history of coastal north-eastern Sardinia (Italy) under changing
Holocene climates and land use. Veget Hist Archaeobot 25:271–
289. https://doi.org/10.1007/s00334-015-0548-5
Beit-Arieh I (1997) The Dead Sea region: an archaeological perspec-
tive. In: Niemi TM, Ben-Avraham Z, Gat J (eds) The Dead Sea:
the lake and its settings. Oxford University Press, New York,
pp249–251
Bennett KD (1996) Determination of the number of zones in a
biostratigraphical sequence. New Phytol 132:155–170. https://
doi.org/10.1111/j.1469-8137.1996.tb04521.x
Beug HJ (2004) Leitfaden der Pollenbestimmung für Mitteleuropa und
angrenzende Gebiete, 2ndedn. Pfeil, München
Birks HJB, Gordon A (1985) Numerical methods in Quaternary pollen
analysis. Academic, London
Birks HJB, Heegaard E (2003) Developments in age-depth modelling
of Holocene stratigraphical sequences. PAGES News 11:7–8
Blaauw M (2010) Methods and code for “classical” age-modelling of
radiocarbon sequences. Quat Geochronol 5:512–518. https://doi.
org/10.1016/j.quageo.2010.01.002
Black E, Brayshaw D, Slingo J, Hoskins B (2011) Future climate of
the Middle East. In: Mithen S, Black E (eds) Water, life and civi-
lisation: climate, environment and society in the Jordan Valley.
Cambridge University Press, Cambridge, pp51–62
Bookman R, Enzel Y, Agnon A, Stein M (2004) Late Holocene lake
levels of the Dead Sea. Bull Geol Soc Am 116:555–571
Borchardt D, Bogardi JJ, Ibisch RB (eds) (2016) Integrated water
resources management: concept, research and implementation.
Springer, Cham
Braun-Blanquet J (1964) Pflanzensoziologie: Grundzüge der Vegeta-
tionskunde, 3rdedn. Springer, Wien
Broshi M, Finkelstein I (1992) The population of Palestine in Iron
Age II. Bull Amer Schools Orient Res 287:47–60. https://doi.
org/10.2307/1357138
Chao CT, Krueger RR (2007) The date palm (Phoenix dactylifera
L.): overview of biology, uses, and cultivation. HortScience
42:1,077–1,082
Chen A, Weisbrod N (2016) Assessment of anthropogenic impact on
the environmental flows of semi-arid watersheds: the case study
of the lower Jordan River. In: Borchardt D, Bogardi JJ, Ibisch RB
(eds) Integrated water resources management: concept, research
and implementation. Springer, Cham, pp59–83
Christensen JH, Krishna Kumar K, Aldrian E etal (2013) Climate
phenomena and their relevance for future regional climate
change. In: Stocker TF, Qin D, Plattner G-K etal (eds) Climate
change 2013: the physical science basis. Contribution of Working
Group I to the Fifth Assessment Report of the Intergovernmental
Panel on Climate Change, pp 1. Cambridge University Press,
Cambridge,pp 217–1,308
Clamer C (2010) Paradies am Meeresrand: Die Palastanlage von Ain
ez-Zara/Kallirrhoë. In: Zangenberg J (ed) Das Tote Meer: Kul-
tur und Geschichte am tiefsten Punkt der Erde. Zabern, Mainz,
pp113–124
Clamer C, Dussart O (1997) Fouilles archéologiques de Aïn Ez-
Zâra/Callirrhoé: villégiature hérodienne. Institut français
d’archéologie du Proche-Orient, Beyrouth
Croudace IW, Rothwell RG (eds) (2015) Micro-XRF studies of sedi-
ment cores. Springer, Dordrecht

Vegetation History and Archaeobotany
1 3
Danin A (1983) Desert vegetation of Israel and Sinai. Cana Publish-
ing House, Jerusalem
Danin A, Orshan G (eds) (1999) Vegetation of Israel, vol 1: desert
and coastal vegetation. Backhuys, Leiden
Davies CP, Fall PL (2001) Modern pollen precipitation from
an elevational transect in central Jordan and its relation-
ship to vegetation. J Biogeogr 28:1,195–1,210. https://doi.
org/10.1046/j.1365-2699.2001.00630.x
Dayan U, Morin E (2006) Flash flood-producing rainstorms over the
Dead Sea: a review. In: Enzel Y, Agnon A, Stein M (eds) New
frontiers in Dead Sea paleoenvironmental research. Geological
Society of America, Boulder, pp53–62
Decker JP (1961) Salt secretion by Tamarix pentandra Pall. For Sci
7:214–217
Dols MW (1977) The “black death” in the Middle East. Princeton
University Press, Princeton
Enzel Y, Bookman R, Sharon D etal (2003) Late Holocene climates
of the Near East deduced from Dead Sea level variations and
modern regional winter rainfall. Quat Res 60:263–273. https://
doi.org/10.1016/j.yqres.2003.07.011
Finsinger W, Tinner W (2005) Minimum count sums for charcoal
concentration estimates in pollen slides: accuracy and potential
errors. Holocene 15:293–297. https://doi.org/10.1191/095968
3605hl808rr
Flexer A, Yellin-Dror A (2009) State of the art. In: Hötzl H, Möller
P, Rosenthal E (eds) The water of the Jordan Valley. Springer,
Berlin, pp15–54
Freiwan M, Kadioğlu M (2007) Climate variability in Jordan. Int J
Climatol 28:69–89. https://doi.org/10.1002/joc.1512
Garfunkel Z, Ben-Avraham Z (1996) The structure of the Dead Sea
basin. Tectonophysics 266:155–176. https://doi.org/10.1016/
S0040-1951(96)00188-6
Ghatasheh NA, Mua’ad M, Faris H (2013) Dead Sea water level
and surface area monitoring using spatial data extrac-
tion from remote sensing images. Int Rev Comput Softw
8(2):892–892,897
Hect A, Gertman I (2003) Dead Sea meteorological climate. In: Nevo
E, Oren A, Wasser SP (eds) Fungal life in the Dead Sea. Gantner,
Haifa, pp69–116
Heim C, Nowaczyk NR, Negendank JF etal (1997) Near East deser-
tification: evidence from the Dead Sea. Naturwissenschaften
84:398–401
Heiri O, Lotter AF, Lemcke G (2001) Loss on ignition as a method for
estimating organic and carbonate content in sediments: reproduc-
ibility and comparability of results. J Paleolimnol 25:101–110
Hillenbrand C (2000) The Crusades: Islamic perspectives. Routledge,
New York
Hirschfeld Y (2006) The archaeology of the Dead Sea valley in the
Late Hellenistic and Early Roman periods. In: Enzel Y, Agnon
A, Stein M (eds) New frontiers in Dead Sea paleoenvironmental
research. Geological Society of America, Boulder, pp215–229
Horowitz A (1979) The quaternary of Israel. Academic, London
Horowitz A (1992) Palynology of arid lands. Elsevier, Amsterdam
Hua Q, Barbetti M, Rakowski AZ (2013) Atmospheric radiocarbon for
the period 1950–2010. Radiocarbon 55:2,059–2,072
Hütteroth W-D, Abdulfattah K (1977) Historical geography of Pales-
tine, Transjordan and Southern Syria in the late 16th century.
Selbstverlag der Fränkischen Geographischen Gesellschaft in
Kommission bei Palm und Enke, Erlangen
Jacobson GL, Bradshaw RHW (1981) The selection of sites for
paleovegetational studies. Quat Res 16:80–96. https://doi.
org/10.1016/0033-5894(81)90129-0
Jain SM, Al-Khayri JM, Johnson DV (eds) (2011) Date palm biotech-
nology. Springer, Dordrecht
Juggins S (1991) Zone V. 1.2. Freeware. DOS Program for the zona-
tion (constrained clustering) of palaeoecological data. In: Steve
Juggins web pages at Newcastle University. https://www.staff.
ncl.ac.uk/stephen.juggins/software.htm. Accessed 5 Oct 2016
Kafle HK, Bruins HJ (2009) Climatic trends in Israel 1970–2002:
warmer and increasing aridity inland. Clim Chang 96:63–77.
https://doi.org/10.1007/s10584-009-9578-2
Klein C, Flohn H (1987) Contributions to the knowledge of the fluctua-
tions of the Dead Sea level. Theor Appl Climatol 38:151–156.
https://doi.org/10.1007/BF00868099
Kołaczek P, Zubek S, Błaszkowski J etal (2013) Erosion or plant
succession—how to interpret the presence of arbuscular myc-
orrhizal fungi (Glomeromycota) spores in pollen profiles col-
lected from mires. Rev Palaeobot Palynol 189:29–37. https://doi.
org/10.1016/j.revpalbo.2012.11.006
Lang G (1994) Quartäre Vegetationsgeschichte Europas: Methoden und
Ergebnisse. Spektrum Akademischer Verlag, Jena
Langgut D, Finkelstein I, Litt T etal (2015) Vegetation and climate
changes during the Bronze and Iron Ages (~ 3600–600 BCE) in
the southern Levant based on palynological records. Radiocarbon
57:217–235. https://doi.org/10.2458/azu_rc.57.18555
Langgut D, Neumann FH, Stein M etal (2014) Dead Sea pollen record
and history of human activity in the Judean Highlands (Israel)
from the Intermediate Bronze into the Iron Ages (ca 2500–500
BCE). Palynology 38:280–302. https://doi.org/10.1080/019161
22.2014.906001
Legendre P, Birks HJB (2012) From classical to canonical ordination.
In: Birks HJB, Lotter AF, Juggins S, Smol JP (eds) Tracking
environmental change using lake sediments. Springer, Dordrecht,
pp201–248
Lelieveld J, Proestos Y, Hadjinicolaou P etal (2016) Strongly increas-
ing heat extremes in the Middle East and North Africa (MENA)
in the 21st century. Clim Chang 137:245–260. https://doi.
org/10.1007/s10584-016-1665-6
Lensky NG, Dvorkin Y, Lyakhovsky V etal (2005) Water, salt, and
energy balances of the Dead Sea. Water Resour Res 41:W12418.
https://doi.org/10.1029/2005WR004084
Lepš J, Šmilauer P (2003) Multivariate analysis of ecological data
using CANOCO. Cambridge University Press, Cambridge
Leroy SAG (2010) Pollen analysis of core DS7-1SC (Dead Sea) show-
ing intertwined effects of climatic change and human activities
in the Late Holocene. J Archaeol Sci 37:306–316. https://doi.
org/10.1016/j.jas.2009.09.042
Leroy SAG, Marco S, Bookman R, Miller CS (2010) Impact of earth-
quakes on agriculture during the Roman-Byzantine period from
pollen records of the Dead Sea laminated sediment. Quat Res
73:191–200. https://doi.org/10.1016/j.yqres.2009.10.003
Litt T, Ohlwein C, Neumann FH etal (2012) Holocene climate vari-
ability in the Levant from the Dead Sea pollen record. Quat Sci
Rev 49:95–105. https://doi.org/10.1016/j.quascirev.2012.06.012
López-Merino L, Leroy SA, Eshel A, Epshtein V, Belmaker R, Book-
man R (2016) Using palynology to re-assess the Dead Sea lami-
nated sediments—indeed varves? Quat Sci Rev 140:49–66
Migowski C, Stein M, Prasad S etal (2006) Holocene climate vari-
ability and cultural evolution in the Near East from the Dead
Sea sedimentary record. Quat Res 66:421–431. https://doi.
org/10.1016/j.yqres.2006.06.010
Moore PD, Webb JA, Collinson ME (1991) Pollen analysis, 2ndedn.
Blackwell, London
Mozingo HN (1987) Shrubs of the Great Basin: a natural history. Uni-
versity of Nevada Press, Reno
Neumann FH, Kagan EJ, Leroy SAG, Baruch U (2010) Vegetation
history and climate fluctuations on a transect along the Dead Sea
west shore and their impact on past societies over the last 3500
years. J Arid Environ 74:756–764. https://doi.org/10.1016/j.
jaridenv.2009.04.015
Neumann FH, Kagan EJ, Schwab MJ, Stein M (2007) Palynology,
sedimentology and palaeoecology of the late Holocene Dead

Vegetation History and Archaeobotany
1 3
Sea. Quat Sci Rev 26:1,476–1,498. https://doi.org/10.1016/j.
quascirev.2007.03.004
Neumann FH, Kagan EJ, Stein M, Agnon A (2009) Assessment of
the effect of earthquake activity on regional vegetation. High-
resolution pollen study of the Ein Feshka section, Holocene Dead
Sea. Rev Palaeobot Palynol 155:42–51. https://doi.org/10.1016/j.
revpalbo.2008.12.016
Odhiambo GO (2017) Water scarcity in the Arabian Peninsula and
socio-economic implications. Appl Water Sci 7:2,479–472,492.
https://doi.org/10.1007/s13201-016-0440-1
Phillips AJL, Alves A, Abdollahzadeh J etal (2013) The Botryospha-
eriaceae: genera and species known from culture. Stud Mycol
76:51–167. https://doi.org/10.3114/sim0021
Rambeau CMC (2010) Palaeoenvironmental reconstruction in the
Southern Levant: synthesis, challenges, recent developments
and perspectives. Philos Trans R Soc A Math Phys Eng Sci
368(5):225–225,248. https://doi.org/10.1098/rsta.2010.0190
Rambeau CMC, Black S (2011) Palaeoenvironments of the southern
Levant 5,000 bp to present: linking the geological and archaeo-
logical records. In: Mithen S, Black E (eds) Water, life and civi-
lisation: climate, environment and society in the Jordan valley.
Cambridge University Press, Cambridge, pp94–104
Rambeau CMC, Gobet E, Grand-Clément E etal. (2015) New methods
for the palaeoenvironmental investigation of arid wetlands, Dead
Sea edge, Jordan. In: Lucke B, Bäumler R, Schmidt M (eds) Soils
and sediments as archives of landscape change. (Erlanger Geog-
raphische Arbeiten 42). Fränkische geographische Gesellschaft,
Erlangen, pp 147–162
Ramoliya PJ, Pandey AN (2003) Soil salinity and water status affect
growth of Phoenix dactylifera seedlings. N Z J Crop Hortic Sci
31:345–353. https://doi.org/10.1080/01140671.2003.9514270
Reille M (1992) Pollen et spores d’Europe et d’Afrique du Nord. Labo-
ratoire de Botanique Historique et Palynologie, Marseille
Reille M (1995) Pollen et spores d’Europe et d’Afrique du Nord. Suppl
1. Laboratoire de Botanique Historique et Palynologie, Marseille
Reille M (1998) Pollen et spores d’Europe et d’Afrique du Nord. Suppl
2. Laboratoire de Botanique Historique et Palynologie, Marseille
Reimer P etal (2013) IntCal13 and Marine13 Radiocarbon age calibra-
tion curves 0–50,000 years cal bp. Radiocarbon 55:1,869–1,887.
https://doi.org/10.2458/azu_js_rc.55.16947
Rimawi O, Salameh E (1988) Hydrochemistry and groundwater sys-
tem of the Zerka Ma’in-Zara thermal field, Jordan. J Hydrol
98:147–163
Robinson SA, Black S, Sellwood BW, Valdes PJ (2006) A review of
palaeoclimates and palaeoenvironments in the Levant and East-
ern Mediterranean from 25,000 to 5000 years bp: setting the
environmental background for the evolution of human civilisa-
tion. Quat Sci Rev 25:1,517–1,541. https://doi.org/10.1016/j.
quascirev.2006.02.006
Salameh E, El-Naser H (2000) Changes in the Dead Sea level
and their impacts on the surrounding groundwater. Acta
Hydrochim Hydrobiol 28:24–33. https://doi.org/10.1002/
(SICI)1521-401X(200001)28:1<24::AID-AHEH24>3.0.CO;2-6
Salameh E, El-Naser H (2009) Retreat of the Dead Sea and its effect
on the surrounding groundwater resources and the stability of its
coastal deposits. In: Hötzl H, Möller P, Rosenthal E (eds) The
water of the Jordan valley: scarcity and deterioration of ground-
water and its impact on the regional development. Springer, Ber-
lin, pp247–264
Salameh E, Udluft P (1985) The hydrodynamic pattern of the central
part of Jordan. Geol Jahrb 38:39–53
Sawarieh A, Hötzl H, Salameh E (2004) Hydrogeology of thermal
waters from Zara-Zarqa Ma’in and Jiza areas, Jordan. In: Chatzi-
petros AA, Pavlides SB (eds) Proceedings of the 5th International
Symposium on Eastern Mediterranean Geology, Thessaloniki,
Greece, 14–20 April 2004, pp1,564–1,567
Sawarieh A, Hötzl H, Salameh E (2009) Hydrogeology and hydro-
chemistry of Wadi Waleh and Wadi Zarqa Ma’in catchment, cen-
tral Jordan. In: Hötzl H, Möller P, Rosenthal E (eds) The water
of the Jordan valley: scarcity and deterioration of groundwater
and its impact on the regional development. Springer, Berlin,
pp361–370
Schiebel V (2013) Vegetation and climate history of the southern
Levant during the last 30,000 years based on palynological inves-
tigation. Universitäts- und Landesbibliothek, Bonn
Seetzen UJ (1854) Reisen durch Syrien, Palästina, Phönicien, die Tran-
sjordan-Länder, Arabia Petraca und Unter-Aegypten. Reimer,
Berlin
Shaer H, Squires V (2015) Halophytic and salt-tolerant feedstuffs:
impacts on nutrition, physiology and reproduction of livestock.
CRC, Boca Raton
Siebert C, Rödiger T, Geyer S etal (2016) Multidisciplinary inves-
tigations of the transboundary Dead Sea basin and its water
resources. In: Borchardt D, Bogardi JJ, Ibisch RB (eds) Inte-
grated water resources management: concept, research and
implementation. Springer, Cham, pp107–127
Smadi MM (2006) Observed abrupt changes in minimum and maxi-
mum temperatures in Jordan in the 20th century. Am J Environ
Sci 2:114–120
Stockmarr JE (1971) Tablets with spores used in absolute pollen analy-
sis. Pollen Spores 13:615–621
Strobel A (1977) Auf der Suche nach Machärus und Kallirrhoe: Selb-
stzeugnisse und Dokumente zu einem geographischen Problem
des 19. Jahrhunderts. Zeitschrift des Deutschen Palästina-Vereins
93:247–267
Stuiver M, Polach HA (1977) Discussion; reporting of C-14 data.
Radiocarbon 19:355–363
Szidat S, Salazar GA, Vogel E etal (2014) 14C Analysis and sample
preparation at the new Bern laboratory for the analysis of radio-
carbon with AMS (LARA). Radiocarbon 56:561–566. https://
doi.org/10.2458/56.17457
Tiner RW (1999) Wetland indicators: a guide to wetland identification,
delineation, classification, and mapping. CRC, Boca Raton
Tinner W, Hu FS (2003) Size parameters, size-class distribution and
area-number relationship of microscopic charcoal: relevance for
fire reconstruction. Holocene 13:499–505. https://doi.org/10.11
91/0959683603hl615rp
UN data (2016) World statistics pocketbook: Jordan. United
Nations statistics division. http://data.un.org/CountryProfile.
aspx?crName=JORDAN. Accessed at 30 Mar 2017
Van Geel B (1986) Application of fungal and algal remains and other
microfossils in palynological analyses. In: Berglund BE (ed)
Handbook of Holocene palaeoecoloy and palaeohydrology.
Wiley, Chichester, pp497–505
Vepraskas MJ, Craft CB (2016) Wetland soils: genesis, hydrology,
landscapes, and classification, 2ndedn. CRC, Boca Raton
Waitzbauer W (2004) Reise durch die Natur Jordaniens. Biologiezen-
trum, Oberösterreichische Landesmuseen, Linz
Weinberger G, Livshitz Y, Givati A etal (2012) The natural water
resources between the Mediterranean Sea and the Jordan River.
Hydrological Service of Israel, Jerusalem
Yechieli Y, Abelson M, Baer G (2016) Sinkhole formation and subsid-
ence along the Dead Sea coast, Israel. Hydrogeol J 24:601–612.
https://doi.org/10.1007/s10040-015-1338-y
Zohary M (1962) Plant life of Palestine: Israel and Jordan. Ronald,
New York
Zohary M, Feinbrun-Dothan N (1966) Flora Palaestina. Israel Acad-
emy of Sciences and Humanities, Jerusalem