This article was downloaded by: [Oded Lipschits]

On: 18 January 2013, At: 06:22

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Palynology
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/tpal20

Fossil pollen reveals the secrets of the Royal Persian

Garden at Ramat Rahel, Jerusalem
a a b a
Dafna Langgut , Yuval Gadot , Naomi Porat & Oded Lipschits
a
Institute of Archaeology, Tel Aviv University, P.O. Box 39040, Tel Aviv, 69978, Israel
b
Geological Survey of Israel, 30 Malkhe Israel St., Jerusalem, 95501, Israel
Accepted author version posted online: 16 Oct 2012.Version of record first published: 17
Jan 2013.

To cite this article: Dafna Langgut , Yuval Gadot , Naomi Porat & Oded Lipschits (2013): Fossil pollen reveals the secrets of
the Royal Persian Garden at Ramat Rahel, Jerusalem, Palynology, DOI:10.1080/01916122.2012.736418

To link to this article: http://dx.doi.org/10.1080/01916122.2012.736418

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 Palynology, 2013
http://dx.doi.org/10.1080/01916122.2012.736418

Fossil pollen reveals the secrets of the Royal Persian Garden at Ramat Rahel, Jerusalem
Dafna Langguta*, Yuval Gadota, Naomi Poratb and Oded Lipschitsa
a
Institute of Archaeology, Tel Aviv University, P.O. Box 39040, Tel Aviv 69978, Israel; bGeological Survey of Israel,
30 Malkhe Israel St., Jerusalem 95501, Israel

The ancient tell (mound) of Ramat Rahel sits on the outskirts of Jerusalem. It features an impressive residency and
palatial garden that flourished during the seventh to fourth centuries BCE, when biblical Judah was under the
hegemony of the Assyrian, Babylonian and Persian empires. Until recently, the garden’s flora has been a mystery,
as standard archaeological procedures were unable to retrieve secure archaeobotanical remains. A unique method
of extracting fossil pollen from ancient plaster has now enabled researchers to reconstruct the exact vegetation
components of this royal Persian garden and for the first time to shed light on the cultural world of the inhabitants
of the residence. The plaster layers and garden are dated archaeologically and by Optically Stimulated
Downloaded by [Oded Lipschits] at 06:22 18 January 2013

Luminescence (OSL) methods to the Persian period (fifth to fourth centuries BCE), and produced evidence of
importation by the ruling Persian authorities of special and highly valued trees to the garden from remote parts of
the empire. The most surprising find, and marking its earliest appearance in the southern Levant, was the citron
(Citrus medica), which later acquired a symbolic-religious role in Judaism. Other imported trees found to have
been grown in the garden are the cedar, birch and Persian walnut. The pollen evidence of these exotic trees in the
Ramat Rahel palatial garden suggests that they were probably brought to flaunt the power of the imperial Persian
administration. Native fruit trees and ornamentals that were also grown there include the fig, grape, olive, willow,
poplar, myrtle and water lily. The identification of the ancient garden’s plant life opens a course for future
research into the symbolic role of flora in palatial gardens. It also offers new opportunities for studying the
mechanism by which native flora was adopted in a particular geographical area and proliferated by humans across
the world.
Keywords: pollen; citron; ancient gardens; Persian Period; Ramat Rahel; exotic trees; Israel

1. Introduction garden in the Levant known to date, prior to the

Pollen grains, the fingerprints of plants, are extremely classical period (Lipschits et al. 2009, 2011). Until
helpful in reconstructing ancient natural vegetation recently, the garden’s flora has been a mystery, as
and climate conditions (Bryant 1990). This paper standard archaeological procedures were unable to
documents how pollen fossils can contribute to the retrieve secure archaeobotanical remains. The soil of
reconstruction of environments fashioned by man that the garden was examined for pollen and phytoliths but
long ago disappeared – in our case, gardens. It is well yielded no results. However, pollen originating from
attested in historical, as well as archaeological, records the flora of the garden may have also been trapped
cross-culturally that artificially planted gardens were within the garden’s different facilities (Fish 1994).
part of many palatial edifices and that usually they Since some of the garden’s walls were covered by layers
displayed the capacity of the palace owner to import of plaster, the study’s hypothesis was that if the plaster
rare and exotic plants and to sustain these plants in had ever been refurbished when the garden was in
their adopted, unnatural habitat (Foster 2004; Conan bloom, the wet plaster surface would have trapped
2007; Avyasaf 2010). pollen. Previous studies have extracted pollen from
The tell (mound) of Ramat Rahel is located 4.5 km Levantine ancient plasters yielding natural vegetation
south of the ancient city of Jerusalem, in the southern components which represent either the surrounding
Levant (Figure 1a–c). The site features a royal vegetation or products of different human activities
residency that flourished during the seventh to fourth within the site (e.g. Weinstein-Evron & Chaim 1999;
centuries BCE (Aharoni 1962, 1964; Lipschits et al. Schoenwetter & Geyer 2000).
2009, 2011). Adjacent to this impressive residency a This research aims to reconstruct for the first time
notable garden was discovered; this is the only palatial the botanical components of a Near Eastern ancient

*Corresponding author. Email: [email protected]

Ó 2013 AASP – The Palynological Society

 2 D. Langgut et al.
Downloaded by [Oded Lipschits] at 06:22 18 January 2013

Figure 1. (a) Map of the southern Levantine region; (b) location map showing the site of Ramat Rahel together with other
locations mentioned in the text; (c) aerial view of Area C1: the garden and some of its facilities. Notes: black stars mark the
location of soil samples taken for Optically Stimulated Luminescence (OSL) dating, and white stars mark the location of samples
taken for pollen analysis from the inner plaster layers of Pool 2. Its inner measurements are ca. 7 6 7 m.

garden based on trapped pollen that originated in (P. lentiscus and P. palaestina; ¼ pistachios),
layers of plaster. In order to determine the exact period Ceratonia siliqua (carob tree) and Pinus hale-
when this palatial garden existed, in addition to the pensis (Aleppo pine). In between the trees or in
archaeological finds, an independent radiometric dat- the clearing semi-shrubs and herbs are common,
ing method based on Optically Stimulated Lumines- forming a batha stage. The dominant families
cence (OSL) was performed. within this formation are Poaceae, Caryophyl-
laceae, Brassicaceae and Asteraceae. Recent
plantations of oaks, Pinus halepensis and Cupre-
2. Research area
ssus sempervirens (Italian Cypress) as well as
2.1. Climate and vegetation exotic trees, Eucalyptus, Casuarina and others,
Ramat Rahel is located in the semi-arid Mediterranean have transformed the landscape of the Judean
climate zone. Winter (November–March) constitutes Hills mainly during the previous century.
the rainy season, with annual rainfall amounting to c. (2) The Irano-Turanian vegetation occurs in the
550 mm (Shahar & Sofer 2011), which rapidly research area mainly on the eastern slopes of
decreases closer to the Judean Desert to the east. the Judean Hills. It is characterised by steppe
Winter is preceded and followed by short transitional vegetation with dwarf shrubs, e.g. Artemisia
seasons – autumn (September–October) and spring herba-alba (white wormwood) and grasses.
(April–May) – in which occasional rains occur, as well
as very hot episodes. Summers are hot and dry (Eshel
2002). 2.2. The Ramat Rahel site and its palatial garden
The Ramat Rahel site is situated on the border The site is located on a prominent hill to the south of
between the Mediterranean vegetation territory to the ancient Jerusalem, close to the main roads connecting
west and the Irano-Turanian vegetation belt to the east Jerusalem to the south and the west and to major
(Zohary 1973; Danin & Plittmann 1987). arable lands. The favourable conditions enjoyed by the
site are hampered by the absence of a permanent
(1) Common trees in the Mediterranean vegetation water source.
territory are Quercus calliprinos (evergreen oak), Archaeological excavations at the Ramat Rahel site
Quercus boisseri and Q. ithaburensis (deci- exposed a royal edifice that was first built at the end of
duous oaks), Olea europaea (olive), Pistacia the eighth century BCE; it went out of use in the third
 Palynology 3

century BCE, after a few phases of rebuilding and The primary function of the pool had been as a water
expansion (Figure 3; Aharoni 1962, 1964; Lipschits distribution device. Its use for gathering water should
et al. 2009, 2011). The edifice is composed of two be viewed as less important. In later periods this pool
ceremonial courtyards, built quarters and a projecting went out of use.
tower surrounded by an artificial garden. To date, this
is the only known palatial garden in the Levant prior
to the classical periods. 3. Materials and methods
The garden was spread over 0.5 ha, possibly even
3.1. Palynology
more. In order to create it, the natural terrain of the
hill had to be reshaped to prepare it for landscaping. 3.1.1. Field sampling
Large amounts of rock, a flint and soft limestone Since Pool 2 is the better preserved of the garden’s
formation, were hewn away from the north, west and facilities, it was chosen to be sampled for pollen
south sides of the tower, flattening the surface around analysis (Figure 1c, Figure 2); two well-preserved
it and creating artificial scarps close to three metres layers of plaster from this pool were sampled. Ten
high that separated the tower from its surroundings. plaster samples were collected: five from Layer I (today
The artificially flattened bedrock was then topped an interior plaster) and five from Layer II (the exterior
with a layer of soil about 50 cm deep brought from plaster, today covered by slaked lime). In order to
elsewhere (Figure 4). Its placement upon the levelled sample Layer I, large pieces of the pool wall were taken
Downloaded by [Oded Lipschits] at 06:22 18 January 2013

limestone surfaces was artificial and the soil was out. Later at the laboratory, the interior plaster (¼
designated by the excavators as ‘garden soil’ Layer I) was gently, and quite easily, separated from
(Lipschits et al. 2009, 2011). Other features that Layer II. The slaked lime that covered the exterior
were added include at least three water pools, three plaster was gently scraped at the laboratory. The
well-built water tunnels, carved stone drains and working hypothesis was that the pollen grains had
possibly an underground reservoir. Pool 2 (Figure 1c been caught during the preparation of the plaster,
and Figure 2) is the better preserved of the garden’s when it was still wet. In order to evaluate this
pools. The walls of the pool are completely plastered hypothesis sub-sampling was undertaken: each plaster
and the plaster layer runs continuously from the walls sample, 0.8 cm in width on average, was divided into
to the floor of the pool, creating rounded edges. Two two samples: the outer part (50.2 cm) which was
openings in pool 2 channeled water into the garden to peeled away using a knife and the second sub-sample
the west through especially esthetic stone drains. which included only inner filling material. However,

Figure 2. A cross-section in the pool. I ¼ Layer I, the interior plaster; II ¼ Layer II, the exterior plaster, covered partly by
slaked lime.
 4 D. Langgut et al.
Downloaded by [Oded Lipschits] at 06:22 18 January 2013

Figure 3. A schematic plan of the Ramat Rahel edifice during the Iron and Persian periods. Phase I was built at the end of the
eighth century BCE; the construction of Phase II is dated to the last third of the seventh century BCE; Phase III is an expansion
of Phase II dating to the fifth to fourth century BCE.

Figure 4. A section in the garden. The artificially flattened white bedrock (right image) at the bottom is topped by a 50-cm deep
layer of soil brought here from elsewhere. The hole (left image) marks the place where one of the Optically Stimulated
Luminescence (OSL) samples was taken.

this subdivision was not always easy to achieve and 3.1.2. Pollen extraction
therefore the two groups (outer layer and inner filling Samples were treated with the following preparation
material) were not sterile and some mixing had to be procedure: One Lycopodium clavatum C. Linnaeus
taken into consideration. tablet (10,679 + 953 spores in average) was added to
 Palynology 5

each sample. Then 10% HCl was gradually added in (Van-Zeist & Bottema 1977). Cereal type pollen, which
order to dissolve the Lycopodium tablet and to remove is distinguished from other grasses by its larger size (at
the calcium carbonates within the sample, until the least 37 mm, e.g. Beug 2004), has thick pollen walls and
reaction between the acid and carbonates ceased. a pronounced annulus around the pore, includes wild
Samples were rinsed with distilled water several times and cultivated cereals in addition to several other Near
until pH 7 was achieved. Next, a density separation Eastern grasses. From regular pollen grain identifica-
was carried out by using zinc bromide (ZnBr2) solution tion one cannot tell whether they are wild or domes-
with a specific gravity of 1.95. After stirring well and ticated cereals (van Zeist et al. 2009). Wild cereals
vortexing, samples were placed in an ultrasonic water (namely Triticum dicoccoides, Hordeum spontaneum,
bath. Sonication was used to loosen fine organic debris H. glaucum, H. bulbosum, Avena sterilis and A. wiestii)
and separate the microscopic particles. Since pro- grow naturally in the Judean Hills (Danin 2004).
longed and high frequency sonication treatment may However, high cereal percentages with no ecological
result in damage to pollen, a frequency of 20 KHz was explanation might be connected to human activity.
used for a maximum of two minutes. After sonication, Results of the terrestrial pollen counts are given in
samples were centrifuged for 20 minutes at 3500 rpm Table 1 and are presented in percentages. The total
(all other steps were followed by only five minutes of number of terrestrial pollen grains counted in each
centrifuging at the same rpm). The floated suspension sample is given at the bottom of each column. The
was then sieved through a 150-mm mesh screen and various pollen types presented in Table 1 are arranged
Downloaded by [Oded Lipschits] at 06:22 18 January 2013

rinsed with distilled water. The sieved material was in two groups: (1) native Mediterranean maquis/forest
treated with acetolysis mixture: nine parts acetic plants; (2) fruit trees and ornamentals. Unrecognisable
anhydride to one part sulphuric acid. Samples were pollen includes damaged or degraded grains that were
incubated at 808C for five minutes. Before and after unidentifiable.
adding the acetolysis mixture the residue was washed Aquatic pollen as well as fern spores were excluded
with glacial acetic acid. After rinsing the samples with from the terrestrial pollen sum and are presented in
water and then ethanol they were mounted in glycerin. absolute numbers in Table 2 together with palyno-
morph concentration values.
Only samples with over 200 pollen grains are
3.1.3. Pollen identification presented in Table 1 and discussed in the text. Several
A light microscope, with magnifications of 2006, 4006 other samples were barren or were found to contain
and 10006 (immersion oil), was used for identifying the too little terrestrial pollen to enable in-depth con-
pollen grains. In each sample all the pollen grains sideration and discussion. Pollen counting stopped
extracted were counted and identified. For pollen only when 500 Lycopodium spores and not less than
identification a comparative reference collection of the eight slides were reached.
Israel pollen flora (Steinhardt National Collections of
Natural History at Tel Aviv University) was used as well
as pollen atlases (Reille 1995, 1998, 1999; Beug 2004). 3.2. OSL dating
The identification of olive was done to the species Luminescence methods date the time that passed since
level since Olea europaea is the only wild-occurring the last event of signal resetting by sunlight or heat
species in Israel (Zohary 1973). This is also the case (Aitken 1998) and they are applied mostly to the
with Pinus halepensis, which is the only naturally- ubiquitous quartz and feldspar minerals. Optically
occurring pine species in the southern Levant (Wein- stimulated luminescence (OSL) dating is based on solar
stein-Evron & Lev-Yadun 2000). The oaks were resetting (or bleaching) of the luminescence signal
distinguished into evergreen and deciduous trees: while during sediment transport and the OSL age gives the
Quercus calliprinos is the only evergreen oak tree in last event of deposition (Wintle 2008). The lumines-
the area. Among the Q. ithaburensis type, some may cence signals grow over time from the natural environ-
have been Q. boissieri which is a deciduous oak species mental ionising radiation, and the laboratory-measured
of the upper mountain zones of the Upper Judean Hills signal intensity is a function of both age and environ-
(Zohary 1973); Q. ithaburensis, on the other hand, is a mental dose rate. The OSL signals are measured in the
tree of lower elevations. However, the two deciduous laboratory and the equivalent dose (De), the amount of
species are palynologically indistinguishable. irradiation that the sample absorbed since it was last
Due to difficulties in the discrimination of the reset, is determined using the single aliquot regenerative
Chenopodiaceae and the genus Amaranthus these two dose (SAR) protocols (Murray & Wintle 2000). The
groups of plants are presented together as Cheno-Am. environmental dose rates (d) are measured in the field
The Family Asteraceae family was divided into two and in the laboratory. The age is calculated from the
pollen types: A. Asteroideae and A. Cichorioideae ratio De/d.
 6 D. Langgut et al.

Table 1. Terrestrial pollen from the pool plaster at Ramat Rahel.

Plaster Layer and Field ID Blooming: Calendar month1

Layer Layer Layer Layer
Ia Ib IIa IIb 1 2 3 4 5 6 7 8 9 10 11 12
Mediterranean Quercus calliprinos type 0.9 2.0 1.2 0.4
maquis/forest Quercus ithaburensis type 0.9 0.0 0.0 0.4
(in %) Pinus halepensis 3.2 25.7 10.5 42.2
Pistacia (Anacardiaceae) 0.0 0.0 0.6 0.0
Phillyrea (Oleaceae) 0.0 3.0 0.6 0.0
Poaceae 1.4 0.0 2.9 0.9
Cerealia type (Poaceae) 4.6 5.9 4.7 0.9
Asteraceae Asteroideae 6.9 0.0 0.0 1.3
Asteraceae Cichorioideae 6.5 5.9 0.6 0.4
Artemisia (Asteraceae) 0.0 0.0 0.0 2.6
Apiaceae 0.9 0.0 0.0 0.0
Cheno-Am 42.4 36.6 5.2 1.3
Caryophyllaceae 4.6 0.0 1.2 0.4
Liliaceae 0.5 2.0 1.2 0.0
Salvia (Lamiaceae) 2.3 0.0 0.0 0.0
Brassicaceae 5.5 0.0 1.7 3.5
Downloaded by [Oded Lipschits] at 06:22 18 January 2013

Plantago lanceolata 0.5 0.0 1.2 0.0

(Plantginaceae)
Rosaceae 1.4 0.0 0.6 1.3
Urticaceae 0.5 3.0 0.6 1.7
Solanaceae 0.0 1.0 1.2 0.4
Rubiaceae 0.5 0.0 0.0 0.0
Fabaceae 0.0 0.0 0.0 0.9
Fruit trees and Salix (Salicaceae) 0.0 0.0 9.3 0.9
ornamentals Populus euphratica 0.0 0.0 0.6 0.9
(in %) Vitis vinifera 0.0 0.0 1.2 0.9
Juglans regia 0.0 0.0 1.2 0.0
Citrus medica 0.0 0.0 32.0 22.2
Ficus carica 0.0 0.0 2.3 0.9
Olea europaea 1.8 2.0 2.9 1.3
Cedrus libani 0.0 0.0 0.6 0.4
Myrtus communis 0.0 0.0 0.6 2.2
Betula (Betulaceae) 0.0 0.0 0.0 1.3
Unrecognisable 14.7 12.9 15.2 10.4
Total terrestrial pollen 100.0 100.0 100.0 100.0

Notes: (1) Green bars represent the blooming season. (2) Percentages represent the occurrence of certain pollen taxa and do not necessarily
point to its ‘‘real’’ relative frequency since the latter is influenced by wind direction, pool wall facing and the distance of the growing taxa from the
pool.

Table 2. Non terrestrial pollen from the pool plaster at Ramat Rahel and palynomorph concentration values.

Plaster layer Layer Ia Layer Ib Layer IIa Layer IIb

Lemna (Lemnaceae) 2 71 101 15
Nymphaea (Nymphaeaceae) 0 2 1 0
Fern spores 1 74 3 20
Total palynomorphs Sum (terrestrial pollen þ aquatic pollen þ fern spores) 227 248 286 229
Lycopodium 280 461 2039 1443
Weight (g) 5.4 0.2 0.7 1.6
Palynomorph concentration values (g/sediment) 1603.2 28,724.4 2,139.8 1,059.2

3.2.1. Field sampling (at a depth of 10–15 cm) and a second one from the
Four samples were collected for OSL dating from two bottom (at a depth of 35–50 cm) of each section
excavation pits in the garden soil. The soil is about (Figure 1c, Figure 4). The sediment was drilled with a
50 cm thick, and one sample was taken from the top 1 inch diameter spiral hand-held auger. To prevent any
 Palynology 7

(years b. 2010)

central age model (Galbraith et al. 1999). Locus - in the excavation; Depth - from the top of the soil; Ext. - external; Cosmic – the cosmic dose rate that takes into account the archaeological
Notes: Grain size of extracted quartz is 74–125 mm. Time-averaged water contents were estimated at 10 + 3%, reflecting seasonal variations. De average and errors were calculated using the
exposure to sunlight, samples were collected under a

180

160
80

80
cover and placed immediately in light-tight bags. An

Age

+
+
+
+
additional sediment sample was collected from each

2320
3560
2420
3040
locality for dose rate assessments.

0.11
0.27
0.14
0.31
3.2.2. Sample preparation and measurements

(Gy)
De
Quartz in the size range of 74–125 mm was extracted

+
+
+
+
using routine laboratory procedures (Porat 2007) in

4.48
6.52
5.27
6.54
appropriate lighting. After wet-sieving to the desired
grain size, carbonates were dissolved with 10%

O-D
(%)
9
26
10
28
hydrochloric acid (HCl). The rinsed and dried sample
was passed through a Frantz magnetic separator to
remove undissolved carbonates, heavy minerals and

aliquots
No. of

16/16
15/17
17/17
16/17
most feldspars (Porat 2006). A 40-minute rinse in

overburden; No of aliquots – the number used for De calculations out of those measured; O-D - over-dispersion, the scatter within sample.
hydrofluoric acid (HF) (42%) was used to dissolve
remaining feldspars and etch the quartz grains,

50
49
43
44
followed by rinsing in 16% HCl to remove any

Dose rate
(mGy/a)
Downloaded by [Oded Lipschits] at 06:22 18 January 2013

+
+
+
+
fluorides which may have precipitated.

1929
1831
2174
2150
OSL measurements were carried out on Risø DA-20
TL/OSL readers (Bøtter-Jensen et al. 2010), equipped
with an integral strontium-90 (90Sr) calibrated beta

(mGy/a)
Cosmic
source with dose rates of *2.5 Gy/min. Stimulation was

172
166
171
164
with blue light-emitting diodes (LED) and detection was
through 7-mm U-340 filters. The SAR protocol
(Murray & Wintle 2000) was used to determine the De

(mGy/a)
Ext. g

740
715
855
862
on 17 aliquots from each sample, and aliquot size was 5
or 3 mm. Each aliquot was irradiated stepwise and
normalised until the natural signal was regenerated. (mGy/a)
Dose response curves were constructed from five dose
Ext. b

1006
939
1135
1112
points, two of which were repeats (a regular recycling
point and an IR-depletion ratio point), and two zero-
dose points. The most representative De + 1s value for
(mGy/a)
Ext. a

each sample was calculated using the central age model

11
11
13
13

(Galbraith et al. 1999). The scatter in the De values in

each sample (noted in Table 3 as ‘over-dispersion’, an
indication of the scatter within the sample beyond that
(ppm)
Ramat Rahel garden soil OSL dating results.

7.1
7.4
8.5
9.6
Th

which would be expected from experimental uncertain-

ties) was used to assess whether the sediment was well
bleached, and the reliability of the age.
(ppm)
2.35
2.15
U

2.7
2.3

Alpha, beta and gamma dose rates were calculated

from the concentrations of the radioactive elements
uranium (U) and thorium (Th) measured by induc-
0.91
0.83
1.00
1.00
(%)
K

tively coupled plasma (ICP)-mass spectroscopy, and

potassium (K) measured by ICP-atomic emission
Depth

spectroscopy. The cosmic dose rate was evaluated

0.13
0.36
0.15
0.50
(m)

from burial depth, including the archaeological over-

burden. Water contents were estimated at 10 + 3%,
reflecting seasonal variation.
Locus
L036
L036
L008
L008

4. Results
Lab Code.
RMR-24
RMR-25
RMR-26
RMR-27

4.1. Palynology
Table 3.

All subsamples of Layers I and II from the inner filling

material contained only a few terrestrial pollen grains
 8 D. Langgut et al.

and some of aquatic origin. In most cases, the 2320 + 80 and 2420 + 80 yr (years before 2010, the
subsamples taken from the outer part of each of the year of measurement). This places the time when the
plaster layers included higher pollen concentration garden was covered by the overlying layers to roughly
values (Table 1). the fourth century BCE – the transition between the
The pollen assemblages of Layer I include only Persian and Hellenistic periods – and it matches
plants belonging to native Mediterranean maquis/ the stratigraphical and typological age established for
forest. Typical wind-pollinated Mediterranean trees, the end of the garden’s use. The samples from the base
common to the Judean Hills, constitute up to 32.7% of of the sections are older and their ages are 3560 + 180
the total pollen sum (calculated from Table 1), e.g.: and 3040 + 160 yr, providing an age for the original
Quercus calliprinos 5 2.0%, Pinus halepensis 5 25.7% garden soil.
and Phillyrea 5 3%. The pollen of herbs and shrubs is
common to the steppe-like open land flora (also called
5. Discussion
batha – a degradation stage of woodland and maquis
under typical Mediterranean climate) but can also be 5.1. The palynological evidence
ruderal (plants that adapted themselves to man-made It is clear that the abundant, well-preserved pollen in
habitats; Zohary 1962). This group of plants is domi- the outer part of the plaster samples reflects the period
nanated by Cheno-Ams (up to 42.4%). during which the plaster was applied. The plasters
The pollen data derived from Layer II are divided were not exposed to pollen from later periods since
Downloaded by [Oded Lipschits] at 06:22 18 January 2013

into two groups (Table 1): the pool went out of use in the early Hellenistic period
(third century BCE) and was covered by a fill of
(1) Typical Mediterranean maquis/forest and slaked lime (Lipschits et al. 2009). Our assumption
batha members: most of those taxa were that most of the pollen grains were trapped during the
identified in the plaster of Layer I; for instance, preparation of the plaster, when it was still wet, is
Mediterranean trees like Quercus calliprinos most probably correct since samples of inner filling
(51.2%) and Pinus halepensis (542.2%). material contained only few terrestrial pollen grains
Cheno-Am appears in significantly lower per- and some of aquatic origin. Those grains probably
centages than in layer I – not exceeding 5.2%. made their way into the plaster via the garden pool
(2) Fruit trees and ornamentals: Olea europaea is water used to prepare the plaster. That can be part of
the only pollen taxon identified in both layers, the explanation of the occurrence of pollen of
reaching values of 2.9% in Layer II. Both hydrophilous plants and of insect-pollinated plants,
samples of Layer II are dominated by the insect- which is not airborne. It may also possibly be the
pollinated tree Citrus medica (up to 32.0%). All reason for the identification of the two clumps (of
other fruit trees and ornamentals appear only in Myrtus and Vitis) and of the surprising evidence of fig
Layer II: Salix (up to 9.3%), Populus euphratica pollen.
(up to 0.9%), Vitis vinifera (up to 1.2%), The two plaster phases, which reflect maintenance
Juglans regia (up to 1.2%), Ficus carica (up to of the garden’s facilities, were most probably prepared
2.3%), Cedrus libani (up to 0.6%), Myrtus in the spring according to the overlap in blossoming
communis (up to 2.2%) and Betula (up to 1.3%). months (April–May; Table 1) of all identified pollen
taxa. The differences in percentages of the same taxa in
The state of preservation of the pollen within the different plaster samples from the same layer (I and II;
different samples was quite similar and unrecognisable Table 1) are due to differences in the aspect of the
pollen percentages therefore appear similar as well plaster sample in relation to dominant wind direction.
(10.4–15.2%; Table 1). The pollen concentration Since the prevailing winds in the spring are mainly of
values ranged from 1059.2 (Sample IIb) to 28,724.4 western and northwestern origin (Eshel 2002; Dayan
(Sample Ib) (Table 2). In two cases in Sample IIb, et al. 2007), more pollen would be expected to be
pollen grains appeared in clumps: a cluster of four deposited in the pool walls facing west and north than
grains of Myrtus communis and another cluster which those facing east and south (Figure 1c). However, the
included three grains of Vitis vinifera. Differences in west-facing wall was not sampled for pollen analysis
the amount of the samples derived from technical since its plaster was not well preserved. The highest
limitations during field sampling (Table 2). pollen concentrations appeared in the four samples
from the north-facing wall, which are presented in
Table 2 and discussed below. Samples from the other
4.2. Chronology two walls were found to contain low pollen concentra-
The OSL ages from the upper samples in the two tions, probably due to wind direction, and are not dis-
sections in the garden soil are the same within errors, cussed further.
 Palynology 9

might explain, for example, the high pine percentages

5.1.1. Native Mediterranean maquis/forest plants characterising Samples Ib and IIb (Table 1). Some of
The natural Mediterranean maquis/forest pollen the herb taxa within the Ramat Rahel pollen assem-
elements extracted from the two plaster phases are blages can also be ruderal pollen indicators, pointing
typical of recent Judean Hills vegetation (Zohary to managed landscape in the immediate vicinity of the
1973) and are considered part of the Late Quaternary site.
Levantine natural flora (Horowitz 1979; Weinstein-
Evron 1983; van-Zeist & Bottema 2009; Langgut et al.
2011). The arboreal vegetation is dominated by pine, 5.1.2. The unique pollen spectrum of plaster Layer II
evergreen oak, deciduous oak, olive and Phillyrea After excluding the native Mediterranean maquis/
while the most widespread herbs and shrubs belong to forest components from the two samples belonging to
the goosefoot, pink, cabbage, cereals and aster the outer plaster layer (Samples IIa and IIb), an
families. interesting pollen assemblage emerged. It comprised
The native Mediterranean maquis/forest plant fruit trees and ornamental plants – some of local
identifications emerging from the palynological data origins while others were imported from far-off lands.
are not only corroborated by modern natural vegeta- Native fruit trees and ornamentals included the
tion cover (Zohary 1973) but also by pollen assem- common fig, grapevine, olive, willow, poplar, myrtle
blages dated to the same period under discussion, and water lily. Imported trees were the citron, cedar of
Downloaded by [Oded Lipschits] at 06:22 18 January 2013

taken from an outcrop in Ein Feshka (Neumann et al. Lebanon, birch and Persian walnut. Yet, since the
2007), an oasis located at the northwest margins of the Cedrus and Betula appear in very low percentages,
Dead Sea, 25 km east of Ramat Rahel (Figure 1a). long-distance transport must be taken into
After excluding the local desert and oasis vegetation consideration.
within the Ein Feshka samples, the palynological data
reflect the Mediterranean pollen grains carried from 5.1.2.1. Imported trees
the Judean Hills by the predominant westerlies Citron (Citrus medica)
(Neumann et al. 2007). The vegetation reconstructed This is the most surprising find among the Ramat
for the Iron Age and the Persian period from pollen at Rahel samples (Plate 1), since it is the earliest botanical
Ein Feshka points towards climatic conditions similar evidence of cultivating citron in the southern Levant.
to those of the present, prior to the increasing humidity Its name, C. medica, suggests Media (Persia), and it is
characterising the Roman and Byzantine periods (Bar- the only citrus crop that was grown in Southwest Asia
Matthews & Ayalon 2004; Neumann et al. 2007). The and the Mediterranean basin in Greek times (Late
Mediterranean elements reflected in the Ein Feshka fourth to late second centuries BCE). Several other
pollen spectrum, such as oaks and pines, are in good citrus crops (e.g. lemon, bitter orange, lime and
agreement with the natural vegetation pollen compo- pummelo) apparently arrived in the Mediterranean
nents at Ramat Rahel. The differences in percentages basin only much later, in early Islamic times (seventh
between the natural vegetation of the two plaster layers century AD; Ramón-Laca 2003). The citron tree seems
and of those from Ein Feshka can be attributed to the to have made its way to Ramat Rahel from India via
distance of the growing vegetation from the pool. That Persia. The fruit, with its characteristic thick aromatic

Plate 1. Citrus medica pollen grains from the ancient garden of Ramat Rahel. Each bar ¼ 10 mm. (a–b) – polar view. (c-d) –
equatorial view.
 10 D. Langgut et al.

rind, was appreciated medicinally (Zohary et al. 2012). 19:23) extol cedar wood from Lebanon and Amanus,
The citron (named etrog in Hebrew, a word of Persian and the Ugarits have left poetic testimony to its
origin) is not mentioned in the Old Testament, and the supremacy (Zohary 1982). The Old Testament also
association between the citron and the Pürıˆ ‘e¨c hädär mentions the use of cedar for prestigious building
(Leviticus 23:40), translated ‘fruit of the goodly tree’, enterprises. Whether historical or not, these references
was made hundreds of years later. During the testify to the high value in which the tree was held. At
Hellenistic period, the citron acquired a symbolic- present, this evergreen conifer tree is widely planted in
religious role in Judaism, especially related to the gardens and parks for ornamentation because of its
Jewish holiday the Feast of Tabernacles. The detailed beauty, robustness and longevity; in the last century
description of Theophrastus (c. 371–287 BCE) in cedars were also planted in Jerusalem as ornamentals.
Historia Plantarum shows that in his day, Citrus Although Cedrus is a heavy pollen-producing tree, it
medica (named by him Persian or Median apple) was appears in relatively low percentages in our pollen
already well established in the Middle East (Zohary assemblages (0.4–0.6%). There are some possible
et al. 2012; but see Biger & Liphschitz 1997). A single explanations for the low frequencies: the Cedrus was
archaeobotanical find suggests an even earlier not at its peak of blooming when the plaster was
arrival: Hjelmqvist (1979) uncovered a few charred maintained; the wind direction, tree location in the
citrus seeds in ca. 3200 BP Bronze Age Hala Sultan garden and the pool wall facing it were not in optimal
Tekke, Cyprus. But these remains have not yet been conditions for the Cedrus pollen to be trapped within
Downloaded by [Oded Lipschits] at 06:22 18 January 2013

directly dated to confirm their antiquity (e.g. by the plaster; long-distance transport from the moun-
radiocarbon dating; Zohary et al. 2012). In addition, tains of Lebanon could also be a possibility.
the presence of seed or fruit remains can point to fruit
importation rather than tree cultivation. Better identi- Persian walnut (Juglans regia)
fied and dated finds have come from three different The Persian walnut tree is not native to the flora of
Roman sites in Egypt (Van der Veen 2001, 2003; Van Israel. It is a traditional nut of Old World agriculture
der Veen & Tabinor 2007). The appearance of fossil which produces beautiful hard timber in addition to its
citron pollen at Ramat Rahel, dated to the fifth to fruit. J. regia grows naturally in the temperate forests
fourth centuries BCE (Plate 1), seems to resolve the of Western Asia (e.g. the Balkans, eastern Turkey and
historical debate (Biger & Liphschitz 1997; Felix 1997) northern Iran). Walnut thrives best in cool, hilly areas
over the first appearance of citron in ancient Israel. It and its cultivation usually benefits from supplementary
should be noted that although Citrus pollen is hardly irrigation in summer (Zohary et al. 2012). Remnants of
ever airborne, it appears in high percentages within the wood of J. regia in Israel are very rare. The earliest
palynological assemblages (up to 32.0%). Therefore it evidence was found in Middle Bronze Age strata at
can be argued that at least one citron tree was located Megiddo (Liphschitz 2000). However, the first appear-
near the pool. ance of walnut pollen in Israel is dated only to the Iron
The Ramat Rahel C. medica pollen grains (Plate 1) Age II (Langgut in press), pointing to intended
are subprolate-spheroidal in shape (as was already cultivation of that tree in the Judean Hills. Based on
reported by others – e.g. Grant et al. 2000; Lippi 2000); Ein Feshka pollen assemblages, J. regia was cultivated
all grains are colporate with a colpus number of four in the region only since the Hellenistic period
and their polar axes frequently measure more than (Neumann et al. 2007). Zohary et al. (2012) claim
32 mm. In general, C. medica grains are larger in size in that walnut cultivation probably started before Roman
comparison to other species in the genus Citrus and times, most likely in northern Iran, northeastern
characterised also by slightly coarser surface ornamen- Turkey and the Caucasus. Its Hebrew name (’e´gôz,
tation (Xianghong 1982). again a word with a Persian origin) appears only once
in the Old Testament (Song of Solomon 6:11). Walnut
Cedar (Cedrus libani) comprises up to 1.4% of the Ramat Rahel pollen
The majestic conifer, the cedar of Lebanon, was assemblages.
never a native forest-tree in Israel. Scattered cedars are
still found in the mountains of Lebanon, probably Birch (Betula spp.)
relics of formerly more widespread distribution. Other The genus Betula contains more then 30 taxa,
remnants are found in northwestern Syria and in the palynologically indistinguishable. It is widespread in
mountain ranges of southern Turkey (Beals 1965). mountainous and temperate zones. Betula’s nearest
From early times, the cedar of Lebanon symbolised occurrence is more than 600 km to the north, at
strength, dignity and grandeur, and was considered the Anatolia (Ercyas Dagi near Kayseri; van Zeist et al.
prince of trees (Zohary 1982). Egyptian and Assyrian 2009). Despite this remoteness, there is some evidence
royal reports as well as the Old Testament (2 Kings of a very long-range transport of Betula pollen (e.g.
 Palynology 11

Hjelmroos 1991). Since Betula appears in low per- cross-pollination (Zohary et al. 2012). Therefore, wild
centages in the Ramat Rahel assemblages (1.3%), grapes account for better pollen distribution. Wild
whether Betula was present in the garden area in the Vitis plants in Israel are a constituent of bank
past or it arrived by long-distance transport is still vegetation (Zohary 1973); hence, the grapevine pollen
questionable. grains found at Ramat Rahel samples are probably
from cultivated plants that were irrigated. The low
5.1.2.2. Native fruit trees and ornamentals pollen dispersal efficiency characterising domesticated
Several fruit trees and ornamentals of local origin were Vitis. Van Zeist et al. (2009) indicates that the grapes
found in the Ramat Rahel garden. The first two taxa, probably grew near the pool. The appearance of Vitis
willow and poplar, inhabit banks of permanent and pollen in a clump of three grains corroborates this
intermittent streams. Therefore, their occurrence assumption.
within the Ramat Rahel pollen assemblages
represents directed planting and controlled irrigation Olive (Olea europaea)
regarding those trees. Olive occurs today in Israel in the Mediterranean
territory both as a cultivated and natural element
Willow (Salix) (Zohary 1973). Before man’s interference with native
Although the willow is insect-pollinated, its pollen vegetation, the wild olive was a component of the
dispersal is relatively good and its appearance in the native Mediterranean Quercus calliprinos–Pistacia pa-
Downloaded by [Oded Lipschits] at 06:22 18 January 2013

pollen spectrum is pronounced (up to 9.3%). Willows laestina association, but in small percentages (Hor-
are very cross-fertile, and numerous hybrids occur, owitz 1979; Weinstein-Evron 1983; van-Zeist &
both naturally and in cultivation. A well-known Bottema 2009; Langgut et al. 2011). This evergreen
ornamental example is the Weeping Willow. Unfortu- tree was one of the first domesticated horticulture
nately willow pollen cannot be identified to the species species in the Old World. Olive has been the most
level (Van-Zeist et al. 2009). Like the citron, willow is prominent, and probably economically the most
one of the ‘Four Species’ used ritually during the important, fruit tree of the Mediterranean basin
Jewish holiday, the Feast of Tabernacles. (Zohary et al. 2012). Over the years, the olive has
been the symbol of peace, wisdom, glory, fertility,
Poplar (Populus) power and purity. Pollen grains (2.8%) of this wind-
Many poplar species are grown as ornamental trees. pollinated tree could possibly have found their way to
They have the advantage of growing very tall, very fast. the plaster from several origins: native or natural
The only poplar species that grows in Israel is P. (escapees) of the Mediterranean maquis/forest char-
euphratica. Although poplar is wind-pollinated, it is acterizing the Judean Hills, from a nearby olive
present in very low values in Ramat Rahel samples (up orchard or directly from the garden.
to 0.9%). One explanation for the low frequency might
be connected to the blooming season: while all other Common fig (Ficus carica)
plants revealed from the palynological spectrum are The Mediterranean fig (F. carica) is the third
spring bloomers (mainly April to June, with some taxa classical fruit crop – after olive and grapevine –
continuing to flower also during the summer; Table 1), associated with the beginning of horticulture in the
P. euphratica blooms from the end of the winter until Mediterranean basin and southwest Asia (Zohary &
the beginning of spring (February to April). In addi- Spiegel-Roy 1975). It is a native fruit tree in the region;
tion, the poplar probably did not grow near the the oldest known fig pips came from the ca. 800,000 BP
immediate vicinity of the pool but in a more remote Acheulean Gesher Benot Ya’akov site, Israel (Mel-
area since it has a very vigorous and invasive root sys- amed et al. 2011). The fig has been part of regular food
tems stretching up to 40 m from the tree; planting production in the Levant since the Early Bronze Age,
poplar close to buildings may result in damaged providing fresh fruit in summer and storable, sugar-
foundations and cracked walls due to its search for rich dry fruit all year round (Zohary et al. 2012). Wild
moisture. common figs grow mainly in the low altitudes of the
Mediterranean vegetation belt, occupying streamsides
Grapevine (Vitis vinifera) but also habitats like rock crevices and gorges.
Grape is one of the most important classical fruits Therefore the figs of Ramat Rahel might have
of the Old World. Nearly all domesticated grapes are depended on irrigation. The common fig is pollinated
monoecious (they produce male and female flowers at by an elaborate symbiosis with a particular species of
different locations on the same plant) and are self- wasp (Blastophaga psenes); the flower is not visible, as
pollinated, while wild grapes are dioecious plants it blooms inside the syconium (Galil & Neeman 1977).
(they have unisexual flowers) with obligatory Therefore, the common fig is usually extremely
 12 D. Langgut et al.

underrepresented in the pollen spectrum. The appear- walls testify that the pool was maintained and renewed
ance of fig pollen within the plaster samples, although over a long period.
in very low values (not exceeding 2.3%) is therefore There are some clues that in the Hellenistic period
surprising. Fig fruits may have fallen and decomposed (third century BCE) the pool was transformed into a
into the water pools that were later used to mix the vat for lime production: the openings of the drains
plaster. were sealed from within by plaster when the final coat
of cement was poured. A second clue is that a thick
Myrtle (Myrtus communis) layer of inert quicklime that filled the pool was found
The Myrtle is a cultivated native evergreen shrub and this also attests to the latest use of the pool. The
and the only member of the Myrtaceae family that quicklime in the pool was probably produced in a
grows naturally in Israel, mainly in the north. The nearby lime kiln to the south of the pool (Figure 1c).
pollen within the plaster could be from the surround- The kiln, which was built after the garden was already
ings of the site or from the garden itself. The latter is abandoned, was dated by pottery found in it to the
more likely because myrtle is insect-pollinated and Hellenistic period.
therefore has low dispersal efficiency. In addition, in Building 824, located farther to the south (Figure
one case it appeared in a clump of four pollen grains, 3), was built into the garden enclosure, cutting into its
which probably indicates the close proximity of myrtle soil and reusing the southeastern corner of the garden’s
shrub(s) to the pool. Because of its deep evergreen enclosure. It is therefore clear that the garden predates
Downloaded by [Oded Lipschits] at 06:22 18 January 2013

color, appealing odor and amenity to clipping to form the building. A small but significant pottery assem-
a hedge, the myrtle was (and is) an indispensable blage was found resting on the floor of the building.
feature in ornamented gardens. Like the citron and The pottery dates to the end of the Persian period
willow, it is one of the ‘Four Species’ used ritually (mid-fourth century BCE) which means the garden has
during the Feast of Tabernacles. Furthermore, in the to date to the Persian and/or the Iron periods. Finally,
Book of Esther, Hadassah is the Hebrew name for the entire garden enclosure, including Pool 2, was
Esther (Aester) which is Persian, because both names covered by an earth fill more than 2 m thick. The latest
have the same meaning, that is, myrtle. pottery sherds and other indicative finds found within
the fill date to the second century BCE (Late
Water lily (Nymphaea) Hellenistic period).
There are two species of water lily that are native to Evidently all archaeological finds point towards
Israel aquatic flora: star lotus (N. nouchali) and white dating the garden and Pool 2 to between the seventh
lotus (N. alba), which are palynologically indistin- and fourth centuries BCE. The OSL dating obtained in
guishable. Both species are valued in water gardening this study verifies the archeological dating and places
and pond decoration because of their spectacular the time when the garden was covered by the overlying
flowers. Like the pollen of the duckweed (Lemna), layers to roughly the fourth century BCE – the
the pollen of the water lily probably made its way into transition between the Persian and Hellenistic periods.
the plaster by the garden water used in plaster Therefore, the plaster was most probably applied to the
preparation. Unlike other ornamentals, these two walls of Pool 2 in the fourth century BCE or earlier.
aquatic plants also appear in the inner plaster of Layer It is worth noting the differences between the upper
I, which means that pond ornamentation preceded the and lower OSL samples from each section (Table 3):
rearrangement of the impressive garden. The occur- the upper samples have low over-dispersion (O-D) of
rence of Nymphaea pollen grains also indicates that the the De values (9–10%) and very similar ages. The
pools within the garden were used not only as a water lower samples have high O-D values (25–28%); they
distribution device (Lipschits et al. 2011) but also for are significantly older, and with substantially larger
aesthetic purposes. The presence of decorated stone- errors on the ages. The low O-D values for the upper
built drains in Pool 2 corroborates this assumption. samples indicate that they were sufficiently well
bleached before they were covered by the overlying
sediment; this can be explained by constant mixing of
5.2. The dating of Pool 2 and the garden the soil by ploughing during the period that the garden
Based on pottery and other related material culture was cultivated, allowing each quartz grain to be
items found in the soil layer used for the garden, as exposed to sunlight over time and ensuring that the
well as fills found under the floors of other architectur- OSL ages are indeed burial ages. The ages of the
al structures of Building Phase II at the site (Figure 3), underlying samples, much older than the archaeologi-
Pool 2 constructions (from which the pollen samples cal context, suggest that they were not bleached when
were taken) date the garden to the end of the seventh the garden soil was brought to the site, or after that
century BCE. The several plaster layers on the pool’s during the cultivation period. The substantially higher
 Palynology 13

O-D values reflect their source, probably mixed The Persian association of the garden can be further
sediments from a stream terrace nearby. linked to the species recognised in the pollen spectrum
that were probably brought from the territory of the
Persian empire: the Persian walnut and the citron –
6. Summary and conclusions which both also preserved their Persian origin in their
Our palynological study offers a window into a past Hebrew name. The birch and the cedar, which are also
that is at least 2400 years old and that not only enables not native members in the study area, might have been
us to scientifically identify ancient flora and visualise a imported from mountainous areas of the provinces of
reconstruction of life much as it was in a Royal Persian the Persian empire; however long-distance pollen
garden; it also gives us a glimpse into the mechanisms transport could also be a possibility for their occurrence
by which native flora was adapted and finally in the Ramat Rahel palynological spectrum.
disseminated throughout the world. Finding trees from remote places at the Ramat Rahel
The two plaster layers were put in place when garden located in Judah, a distant province of the
typical Mediterranean vegetation existed in the Judean Persian empire, may help to illustrate the mechanism
Hills with the occurrence of some ruderal plants within through which elite gardens contributed to the spread of
the immediate vicinity of the site, representing human flora across the globe (Foster 1998). The desire to present
interference on the nearby vegetation. Climate condi- in gardens trees from the distant parts of the empire was
tions at that time were similar to those of today. When part of the royal display of power and propaganda.
Downloaded by [Oded Lipschits] at 06:22 18 January 2013

Layer I plaster was formed, only Mediterranean Some of those trees, first brought for their symbolic
vegetation existed in the surrounding areas of the value, became part of the local religious symbolic world
site; at the time that the plaster of Layer II was created, and economic horticulture, as in the case of the citron.
the royal garden had already been established. The well-watered imperial Persian garden at Ramat
Various trees identified in the palynological spec- Rahel must have left a lasting impression on viewers in
trum (e.g. willow, poplar) naturally inhabit river this relatively arid environment. Its imported trees
banks. In order to grow those trees, as well as some from far-off lands, aromatic plants and impressive fruit
of the fruit trees at Ramat Rahel, an intensive and trees, together with its aesthetic architectural features,
controlled irrigation system was most probably re- symbolised the power and affluence of the Persian-
quired. The site of Ramat Rahel has no natural water period rulers who reigned at the Ramat Rahel palace.
source and is completely dependent on the collection of
rain water. The sophisticated water installations that Acknowledgements
make up a complex water collection and storing system This study was funded by the European Research Council
revealed during the excavations at the site (Lipschits under the European Community’s Seventh Framework Pro-
et al. 2011) testify to the high investment needed in gram (FP7/2007-2013)/ERC Grant agreement no. 229418,
and by a grant of ‘Early Israel’ (New Horizons Project) at
order to sustain the garden and its flora. Tel Aviv University. We are grateful to I. Finkelstein, S. Lev-
Reconstruction of the layout and landscaping of the Yadun and two anonymous reviewers for their helpful
garden is within our reach. Undoubtedly, spectacular remarks, to A. Sandler for soil identification and E. Weiss,
aquatic plants like the water lily floated in the garden’s M. Pollak, M. Kitin and S. Ben Dor Evian for their
pools. The soil located around the pools was probably assistance.
used for shrubs or small trees like the myrtle and the
grapevine. The citron and other insect-pollinated plants Author biographies
were probably planted in the near vicinity of the pool.
Some of the great trees (e.g. poplar with its invasive DAFNA LANGGUT received her Ph.D. in
archaeology in 2008 from The University of
root system) were probably planted farther away, at the Haifa. She is a researcher at the Sonia and
outer perimeter of the garden. Marco Nadler Institute of Archaeology, Tel
The garden was lush with local fruit trees, local Aviv University. Her dissertation dealt with
ornamentals and also exotic trees brought from afar. vegetation and climate reconstruction based
This choice of flora to be planted in the garden fits our on fossilised palynomorphes (pollen, spores
and dinoflagellates) extracted from eastern Mediterranean
knowledge of Persian royal gardens and the concept of marine cores of the last 90,000 years. Recently she completed
pardesu (paradise ¼ garden; Foster 1998, 2004; Hunt her postdoctoral research, at the Department of Archae-
2011). Accordingly, royal palatial gardens of the ology and Near Eastern Cultures at Tel Aviv University as
Achamenid empire had to be planned artificially and part of the project, ‘‘Reconstructing Ancient (Biblical) Israel:
their cultivation would have been both practical and The Exact and Life Sciences Perspective’’ (funded by the
European Research Council). Within this research she
aesthetic. The Ramat Rahel garden as observed studied in high resolution the past vegetation of ancient
through the pollen spectra fulfills all these Israel during the Bronze and Iron Ages and the past
requirements. relationship between humans and the environment, such as
 14 D. Langgut et al.

the onset of agriculture, de-forestation and settlement Aharoni Y. 1964. Excavations at Ramat Rahel Seasons
history. Dr Langgut also extracts botanical remains from 1961–1962. Rome: Universita Centro di studi semitici.
archeological sites and deals with utilization patterns for Aitken MJ. 1998. An introduction to optical dating. Oxford:
living spaces, diet, plant usage, agricultural practices, plant Oxford University Press.
importation, ancient gardens and seasonality of site occupa- Avyasaf RS. 2010. Gardens at a crossroads: the Influence of
tion. In addition, she is also a curator of the Archaeobotany Persian and Egyptian gardens on the Hellenistic royal
National Collections of Natural History at Tel Aviv gardens of Judea. Bollet Archeol. 1:27–37.
University. Bar-Matthews M, Ayalon A. 2004. Speleothems as palaeo-
climate indicators: a case study from Soreq cave located in
the Eastern Mediterranean region, Israel. In: Battarbee R,
YUVAL GADOT obtained his Ph.D. in Gasse F, Stickley CE, editors. Past climate variability
archaeology in 2004 from Tel Aviv Uni- through Europe and Africa. Dordrecht: Springer; p. 363–
versity. He is a research associate at the 391.
Sonia and Marco Nadler Institute of Beals EW. 1965. The remnant Cedar forest of Lebanon. J
Archaeology, Tel Aviv University. Since Ecol. 53:679–694.
2011 Dr Gadot has been the director of a Beug HJ. 2004. Leitfaden der Pollenbestimmung für Mitte-
new research project at the Rephaim valley, leuropa und angrenzende Gebiete. München: Dr. Frie-
the bread basket of ancient Jerusalem, designed at recon- drich Pfeil.
structing human strategies for the exploitation of the valley Biger G, Liphschitz N. 1997. Is the citron the Biblical ‘‘Hadar
within their historical setting. Dr Gadot is the field director Fruit’’?. Beit Mikra 148:28–33 (in Hebrew).
of the Ramat Rahel excavations and the archaeological Bøtter-Jensen L, Thomsen KJ, Jain M. 2010. Review of
Downloaded by [Oded Lipschits] at 06:22 18 January 2013

director of Reconstructing Ancient (Biblical) Israel: The optically stimulated luminescence (OSL) instrumental
Exact and Life Sciences Perspective, a project funded by the developments for retrospective dosimetry. Radiat Meas.
European Research Council (ERC). He is also the co- 45:253–257.
director of the community-based archaeological project at Bryant VM. 1990. Pollen: nature’s fingerprints of plants.
the city of Lod. Yearbook of Science and the Future. Chicago: Encyclo-
pedia Britannica.
Conan M. 2007. Middle East garden traditions: unity and
NAOMI PORAT received her Ph.D. in geoarchaeology from diversity. Washington: Dumbarton Oaks.
The Hebrew University in 1989. She is a senior research Danin A. 2004. Distribution atlas of plants in the Flora
scientist at the Geological Survey of Israel. Her expertise is in Palaestina area. Jerusalem: Israel Academy of Sciences
luminescence dating and Quaternary geochronology. In 1994 and Humanities.
she established the luminescence dating laboratory at the GSI Danin A, Plittmann U. 1987. Revision of the plant
and since then has published a large number of papers in that geographical territories of Israel and Sinai. Plant System
field. Evol. 156:S.43–S.53.
Dayan U, Ziv B, Shoob T, Enzel Y. 2007. Suspended dust
ODED LIPSCHITS obtained his Ph.D. in over south-eastern Mediterranean and its relation to
Jewish History from Tel Aviv University in atmospheric circulations. Int J Climatol. 24(8):1001–
1997. He is a Professor of Jewish History at 1011.
the Jacob M. Alkow Department of Ar- Eshel G. 2002. Mediterranean climates. Israel J Earth Sci.
chaeology and Ancient Near Eastern Cul- 51:157–168.
tures, the Director of the Sonia and Marco Felix Y. 1997. Citrus fruit – the Citron. Beit Mikra 148:288–
Nadler Institute of Archaeology and the 292 (in Hebrew).
Head of the ‘‘Ancient Israel Studies Program’’ at Tel Aviv Fish SK. 1994. Archaeological palynology of gardens and
University. Since 2004 he has been the director of the fields. In: Miller NF, Gleason KL, editors. The archae-
renewed excavations project at Ramat Rahel, Jerusalem, ology of garden and field. Philadelphia: University of
Israel (together with Prof. Manfred Oeming from the Pennsylvania Press; p. 44–69.
University of Heidelberg), and since 2011 he has been the Foster KP. 1998. Gardens of Eden: exotic flora and fauna in
director of the Lautenschläger Azekah Expedition (together the Ancient Near East. Yale Forest Environ Studies Bull.
with Prof. Manfred Oeming and Dr Yuval Gadot). Prof. 103:320–329.
Lipschits’ research focuses on Judah under the rule of the Foster KP. 2004. The Hanging gardens of Ninveh. Iraq
Assyrian, Babylonian and Persian Empires (see his 2005 66:207–220.
book – The Fall and Rise of Jerusalem: The History of Judah Galil J, Ne’eman G. 1977. Pollen transfer and pollination in
under Babylonian Rule, Winona Lake), as well as on the the common fig (Ficus carica L.). New Phytol. 79:163–
administration of Judah during this period (and see his 2011 171.
book together with Prof. David S. Vanderhooft – Yehud Galbraith RF, Roberts RG, Laslett GM, Yoshida H, Olley
Stamp Impressions: A Corpus of Inscribed Stamp Impressions JM. 1999. Optical dating of single and multiple grains of
from the Persian and Hellenistic Periods in Judah, Winona rock from Jinmium Rock Shelter, Northern Australia,
Lake). Part 1: Experimental design and statistical models.
Archaeometry 41:339–364.
Grant M, Blackmore S, Morton C. 2000. Pollen morphology of
the subfamily Aurantioideae (Rutaceae). Grana 39:8–20.
References Hjelmqvist H. 1979. Some economic plants and weeds from
Aharoni Y. 1962. Excavations at Ramat Rahel Seasons the Bronze Age of Cyprus. Studies Med Archeol. 45:110–
1959–1960. Rome: Universita Centro di studi semitici. 117.
 Palynology 15

Hjelmroos M. 1991. Evidence of long-distance transport of Reille M. 1999. Pollen et Spores d’Europe et d’Afrique du
Betula pollen. Grana 30:215–228. Nord. Seconde édition. Marseille: Laboratoire de Bota-
Horowitz A. 1979. The Quaternary of Israel. New York: nique Historique et Palynologie.
Academic Press. Schoenwetter J, Geyer PS. 2000. Implications of archae-
Hunt P. 2011. Persian Paradise Gardens: Eden and Beyond ological palynology at Bethsaida, Israel. J Field Archae-
as Chahar Bagh. 29.7.2011. Electrum Magazine. Avail- ol. 27:63–73.
able from: http://www.electrummagazine.com/2011/07/ Shahar A, Sofer T. 2011. The new atlas of Israel – The
paradise-gardens-of-persia-eden-and-beyond-as-chahar- national atlas. Jerusalem: Survey of Israel and Hebrew
bagh/ University of Jerusalem.
Langgut D. (in press) Southern Levant pollen record, Palaeo- Van der Veen M. 2001. The botanical evidence. In: Maxfield
climate and human impact from the Late Bronze Age to VA, Peacock DPS, editors. Survey and excavations at
the Persian period. In: Proceedings of the 8th Interna- Mons Claudianus 1987–1993, Volume 2. Institut Fran-
tional Congress on the Archaeology of the Ancient Near çais d’Archéologie Orientale du Caire; p. 174–247.
East (ICAANE 2012). Warsaw: Harrassowitz Verlag. Van der Veen M. 2003. When is food a luxury?. World
Langgut D, Almogi-Labin A, Bar-Matthews M, Weinstein- Archaeol. 34:405–427.
Evron M. 2011. Vegetation and climate changes in the Van der Veen M, Tabinor H. 2007. Food, fodder and fuel at
south eastern Mediterranean during the last glacial- Mons Porphyrites: The botanical evidence. In: Maxfield
interglacial cycle: new marine pollen core evidence for the VA, Peacock DPS, editors. The Roman imperial quarries,
last 86 ka. Quat Sci Rev. 30:3960–3972. survey and excavation at Mons Porphyrites 1994–1998,
Liphschitz N. 2007. Timber in ancient Israel: dendroarch- Volume 2. London: Egypt Exploration Society; p. 83–142.
aeology and dendrochronology, Monograph series 26. Van-Zeist W, Bottema S. 1977. Palynological investigations
Downloaded by [Oded Lipschits] at 06:22 18 January 2013

Tel Aviv: Tel Aviv University. in Western Iran. Palaeohistoria 109:19–95.

Liphschitz N. 2000. Archaeobotanical finds. In: Finkelstein I, Van-Zeist W, Bottema S. 2009. A palynological study of the
Ussishkin D, Halpem B, editors. Megiddo III: The 1992- Acheulian site of Gesher Benot Ya’aqov, Israel. Vegetat
1996 Seasons. Monograph series 18. Tel Aviv University, Hist Archaeobot. 18:105–121
Tel Aviv. Van-Zeist W, Baruch U, Bottema S. 2009. Holocene
Lippi MM. 2000. The garden of the ‘‘Casa delle Nozze di Palaeoecology of the Hula Area, northeastern Israel.
Ercole ed Ebe’’ in Pompeii (Italy): Palynological inves- In: Kaptijn E, Petit LP, editors. A timeless vale,
tigations. Plant Biosyst. 134:205–211. archaeological and related essays on the Jordan Valley.
Lipschits O, Oeming M, Gadot Y, Arubas B. 2009. The 2006 Leiden: Leiden University Press; p. 29–64.
and 2007 excavation seasons in Ramat Rahel. Israel Weinstein-Evron M. 1983. The paleoecology of the Early
Explorat J. 59:1–20. Würm in the Hula Basin, Israel. Paléorient 9(1):5–19.
Lipschits O, Gadot Y, Arubas B, Oeming M. 2011. Palace Weinstein-Evron M, Chaim, S. 1999. Palynological investi-
and village, paradise and oblivion. Unraveling the riddles gation in Sumaqa: 1995-1996. In: Dar S, editor. A
of Ramat Rahel. Near Eastern Archaeol. 74:2–49. Roman and Byzantine Jewish village on Mount Carmel,
Melamed Y, Kislev M, Weiss E, Simchoni O. 2011. Israel. Oxford: BAR International; p. 365–368.
Extinction of water plants in the Hula Valley: Signs for Weinstein-Evron M, Lev-Yadun S. 2000. Palaeoecology of
global warming? J Human Evol. 60:320–327. Pinus halepensis in Israel in the light of archaeobotanical
Murray AS, Wintle AG. 2000. Luminescence dating of data. In: Ne‘eman G, Trabaud L, editors. Ecology,
quartz using an improved single-aliquot regenerative- biogeography and management of Pinus halepensis and
dose protocol. Radiat Meas. 32(1):57–73. Pinus brutia forest ecosystems in the Mediterranean
Neumann FH, Kagan EJ, Schwab MJ, Stein M. 2007. Basin. Leiden: Backhuys Publishers; p. 119–130.
Palynology, sedimentology and palaeoecology of the Wintle AG. 2008. Luminescence dating: where it has been
late Holocene Dead Sea. Quatern Sci Rev. 26: 1476– and where it is boing. Boreas 37:471–482.
1498. Xianghong YYKYZ. 1982. Studies on the pollen morphol-
Porat N. 2006. Use of magnetic separation for purifying ogy of Citrus plants. Scientia Agricultura Sin. 112: 88–93.
quartz for luminescence dating. Ancient TL 24:33–36. Zohary D, Hopf M, Weiss E. 2012. Domestication of plants
Porat N. 2007. Analytical procedures in the luminescence in the Old World. 4th ed. Oxford: Oxford University
dating laboratory. Geological Survey of Israel Technical Press.
Report TRGSI/2/2002. Jerusalem: Geological Survey of Zohary M. 1962. Plant life of Palestine (Israel and Jordan).
Israel (in Hebrew). New York: Ronald Press.
Ramón-Laca L. 2003. The introduction of cultivated citrus Zohary M. 1973. Geobotanical foundations of the Middle
to Europe via Northern Africa and the Iberian Peninsula. East. Stuttgart/Amsterdam: Gustav Fischer/Swet and
Econ Bot. 57:502–514. Zeitlinger.
Reille M. 1995. Pollen et Spores d’Europe et d’Afrique du Zohary M. 1982. Plants of the Bible. Cambridge: Cambridge
Nord. Suppl. 1. Marseille: Laboratoire de Botanique University Press.
Historique et Palynologie. Zohary M, Spiegel-Roy P. 1975. Beginning of fruit growing
Reille M. 1998. Pollen et Spores d’Europe et d’Afrique du in the old world. Science 187: 319–327.
Nord. Suppl. 2. Marseille: Laboratoire de Botanique
Historique et Palynologie.