Journal of Archaeological Science 33 (2006) 1514e1535

http://www.elsevier.com/locate/jas

Geoarchaeology of Sidon’s ancient harbours, Phoenicia

Nick Marriner a,*, Christophe Morhange a, Claude Doumet-Serhal b
a
CNRS CEREGE UMR 6635, Université Aix-Marseille, BP 80 Europôle de l’Arbois, 13545 Aix-en-Provence, France
b
British Museum, LBFNM, 11 Canning Place, London W85 AD, UK
Received 8 November 2005; received in revised form 2 February 2006; accepted 7 February 2006

Abstract

Geoarchaeological data from Sidon’s ancient harbour areas elucidate six evolutionary phases since the Bronze Age. (1) At the time of Sidon’s
foundation, during the third millennium BC, medium sand facies show the city’s northern and southern pocket beaches to have served as proto-
harbours for Middle to Late Bronze Age societies. (2) Towards the end of the Late Bronze Age and the Early Iron Age, expanding international
trade prompted coastal populations into modifying these natural anchorages. In Sidon’s northern harbour, transition from shelly to fine-grained
sands is the earliest granulometric manifestation of human coastal modification. The lee of Zire island was also exploited as a deep-water an-
chorage, or outer harbour, at this time. (3) Although localised sediments evoke developed port infrastructure during the Phoenician and Persian
periods, high-resolution reconstruction of the northern harbour’s Iron Age history is problematic given repeated dredging practices during the
Roman and Byzantine periods. (4) Fine-grained silts and sands in the northern harbour are coeval with advanced Roman engineering works,
significantly deforming the coastal landscape. Bio- and lithostratigraphical data attest a leaky lagoon type environment, indicative of a well-
protected port. (5) The technological apogee of Sidon’s northern harbour is recorded during the late Roman and Byzantine periods, translated
stratigraphically by a plastic clays unit and brackish lagoon fauna. (6) A final semi-abandonment phase, comprising coarse sand facies, concurs
silting up and a 100e150 m progradation of the port coastline after the seventh century AD. We advance three hypotheses to explain these strati-
graphic data, namely cultural, tectonic and tsunamogenic. Finally, our results are compared and contrasted with research undertaken in Sidon’s
sister harbour, Tyre.
Ó 2006 Elsevier Ltd. All rights reserved.

Keywords: Geoarchaeology; Coastal geomorphology; Ancient harbour; Stratigraphy; Phoenicia; Lebanon

1. Introduction
The SidoneDakerman area chronicles a long history of
human occupation stretching back to the Neolithic [67] (see
Fig. 1). Canaan’s oldest city according to Genesis, the tell
occupies a modest rocky promontory that overlooks a partially
drowned sandstone ridge and two marine embayments [17].
During the Iron Age, this geomorphological endowment al-
lowed Sidon to evolve into one of Phoenicia’s key city-states,
producing and transiting wealthy commodities to trading part-
ners in Assyria, Egypt, Cyprus and the Aegean. This trading
ascendancy is corroborated by the Old Testament’s use of the
term Sidonian to encapsulate all Phoenicians. Sidon enjoyed
its apogee during the sixth to fifth centuries BC, at which
time it superseded Tyre as Phoenicia’s principal naval base.
Although Sidon has a long history of archaeological re-
search [13e15,20e24,64] the ancient city had, until very re-
cently, never been systematically explored. In light of the
difficult geopolitical context, it was only in 1998 that the Leb-
anese Directorate General of Antiquities authorised the British
Museum to begin systematic excavations of the ancient tell
[16]. Seven years on, a continuous stratigraphy spanning the
third millennium BC through to the Iron Age has been estab-
lished for the city [17].
2. Research aims

* Corresponding author. Tel.: þ33 4 4297 1584; fax: þ33 4 4297 1549. In tandem with the terrestrial excavations, 15 cores were
E-mail address: nick.marriner@wanadoo.fr (N. Marriner). drilled in and around Sidon’s ancient port areas, with three

0305-4403/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jas.2006.02.004
 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535 1515

SEA
CASTLE

W E

OUTER
HARBOUR
Wind rose 1-15
r
S knots >15 ate
knots akw
Bre IV
Inner II
mole

INNER
HARBOUR

IX
VI V
I
XV
III VII

Aeolianite
XII
ridge

XIV BRITISH MUSEUM

EXCAVATION

XIII

X
CASTLE

DOMINANT
LONGSHORE
CURRENT
CRIQUE
RONDE
OPEN
SOUTHERN
HARBOUR XI
0 100 200 m

SANDSTONE
RIDGE I
CORE

25˚ 30˚ 35˚

20˚ 40˚N

VIII 35˚N

Cyprus Sidon
Eastern Mediterranean Sea Tyre
30˚N
N
Nile

CHALCOLITHIC SITE 0 200 km

OF DAKERMAN

Fig. 1. Sidon’s ancient harbour areas and location of cores.

main objectives: (1) to elucidate the evolution of the city’s
maritime façade and investigate its coastal palaeogeography
[26,49]; (2) to compare and contrast these data with Sidon’s
sister harbour, Tyre [45,47]; and (3) to investigate human
coastal impacts, and more specifically the problem of acceler-
ated coastal sedimentation. Silting of the Mediterranean’s
ancient ports is a recurrent theme in coastal geoarchaeology,
playing a significant role in littoral progradation and human ex-
ploitation of the anchorages [3,4,50,58,59]. Ancient societies
strived permanently with the silting problem, and indeed in
areas of high sediment supply it was a constant endeavour to
maintain a viable draught depth [46]. From a geoarchaeological
standpoint, the silting provides a multiplicity of research possi-
bilities, not least because the fine-grained sediments and high
water table anoxically preserve otherwise perishable artefacts,
but also the port sediments are a high-resolution sedimentary
archive, recording much of the site’s maritime and occupation
histories.
3. Geomorphological context of Sidon’s maritime façade
Sidon’s coastal plain runs from the Litani river in the south,
northwards towards the Awali (Fig. 2). This low-lying topog-
raphy, up to 2 km wide in places, comprises a rectilinear
1516 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535

Fig. 2. Sidon’s coastal bathymetry.

coastline. In the Sidon area, a series of faults has oriented the
valleys and talwegs NWeSE [18,19,69]. The most important
regional watercourses include the Litani, with headwaters in
the Beqaa valley, and the Awali, which flows from the Jurassic
anticlinal of Barouk-Niha. These watercourses alone transit
w280
106/m3 and 130
106/m3 of sediment per year.
Sidon’s coastal physiography makes it an ideal location for
the establishment of three natural anchorage havens. Two
pocket beaches lie leeward of a Quaternary sandstone ridge,
partially drowned by the Holocene marine transgression
(Fig. 3). To the south of the ancient city this ridge has been
breached by the sea to form a large semi-circular embayment.
 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535
Fig. 3. Sidon and Zire (from [31]). In the foreground, Sidon’s outer harbour lies in the shadow zone of Zire island. The promontory of Sidon separates two coves,
the northern harbour and Poidebard’s Crique Ronde.
Named the Egyptian harbour by Renan [64] and later the
Crique Ronde by Poidebard and Lauffray [57], this coastal
zone presently comprises a sandy beach. Whether or not it
was ever artificially protected by harbourworks has never
been unequivocally demonstrated [57], a question we eluci-
date later in this paper.
North-west of the promontory lies a second bay, protected
from the open sea by a prominent sandstone ridge. Five hun-
dred and eighty meters in length, this coastal ridge shields
a shallow basin still used to this day; a Medieval sea castle,
built upon a small islet, closes the northern portion of the ba-
sin. This northern harbour, the centre of Sidon’s economic and
military activity in antiquity, is mentioned for the first time by
Pseudo-Scylax who describes it as a closed harbour. Much of
Poidebard and Lauffray’s work was centred around this area
where they identified a series of juxtaposed harbourworks.
From their research emerged the vestiges of a closed ancient
port comprising: (1) a reinforced sandstone ridge; and (2) an
artificial inner harbour mole, perpendicular to the ridge, and
separating two basins.
A third harbour area, the offshore island of Zire, is a unique
feature of the Sidonian coastline (Fig. 3). First described by
Renan [64], it was not until the work of Poidebard and
Lauffray [57] that a preliminary plan of the island, with its
quarries and harbourworks, was drawn up. They identified
a double seawall sheltering a series of quarries and harbour
quays on its leeward side. In 1973, underwater surveys
by Honor Frost complemented her predecessors’ work,
1517
uncovering a collapsed jetty and numerous scattered maso-
nary blocks on the sea bottom in proximity to the island
[29]. She concluded that the island had not only served as
a quarry and harbour but also supported a number of con-
structions. Carayon [11] undertook the most recent archaeo-
logical work of note, in which he describes six quarry
zones, detailing the cartography of Poidebard and Lauffray
[57]. During our field investigations we surveyed and dated
an uplifted marine notch (þ50 cm) on these quarry faces, per-
taining to a short-lived sea-level oscillation around 2210  50
BP [44]. These data are in contrast with Tyre, where submer-
gence of w3 m is recorded since late antiquity by coastal
stratigraphy, submerged urban quarters and harbourworks
[25,45,47].
4. Methods and data acquisition
A series of 15 cores was drilled around the two marine em-
bayments (Figs. 1 and 2). Mechanised corers, with 200 cm by
10 cm chambers, were deployed to investigate the coastal stra-
tigraphy and accurately reconstruct the evolution of the an-
cient city’s maritime façade over the past 6000 years. All
cores were GPS levelled (10 cm) and benchmarked relative
to present biological sea level (i.e. the summit of the subtidal
zone, 5 cm [40]). These data derive from the upper limit of
contemporary Balanus populations living on the modern har-
bour quay faces [40].
1518 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535

BHI
+ 1.7

+1
Infill Modal Mode
Sediment Sand Grain size grain Sorting in

Unit
(in mm) index Skewness microns
texture texture histograms (Weydert, 1973)

-0.4
-0.2

0.2
0.2
0.4
0.6
0.8

1.2
25 50 75 25 50 75 0 0.2 0.4 0.6 0 -0.6 0 0.4

1
Biological mean
0 20 I-29
sea level 15

%
10
Sands 5
0
µm
Islamic
shards
20
15
I-26 A

%
10
-1 5
0
20
I-24
15
Muddy sands

%
10
5
0
µm

-2
Depth (m)

20 I-13
15
B1
%

10
Plastic clays 5
0
µm

-3

20 I-11
15
%

10
Silty sands 5
0
and clays 20 I-8 B2
15
-4
%

10
5
0
4931 ± 62 BP
3490 - 3090 BC

-5

Shelly sands C
20 I-3
15
%

10
5
0
-6

20 I-2
Silts and 15
%

10
sands 5 D
0
-7 Transgressive
pebbles
Substratum
Gravels Coarse sands [2mm;500µm]

Sands Medium sands [500µm;200µm]

Silts and clays Fine sands [200µm;50µm]

Fig. 4. Sedimentology of core BH I (northern harbour).

In-depth discussion of the techniques employed can be 5. Results

reviewed in Marriner et al. [45]. These include multi-proxy
litho- (sedimentology and grain size analyses) and biostrati-
graphical (marine molluscan faunas and ostracods) lines of in-
vestigation. The robustness of such techniques to reconcile
complex geoarchaeological questions has been demonstrated
at a number of sites around the circum Mediterranean
[10,39,48,63]. Radiocarbon datings provide a working chrono-
stratigraphic framework. Material included marine shells,
wood fragments, charcoal and seeds, calibrated using the Oxcal
program (see Table 1). All marine material has been corrected
for reservoir effects.
Detailed descriptions of the litho- and biostratigraphical
data from Sidon’s closed northern harbour and the Crique
Ronde follows.
5.1. Northern harbour facies
5.1.1. Unit DdTransgressive contact
Unit D comprises a marine pebble unit which overlies the
sandstone substratum and marks the Holocene marine trans-
gression of the harbour area (Figs. 4, 7 and 10). Many of the
 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535 1519

Table 1
Radiocarbon determinations and calibration
Sample Depth Lab code Material dated 13C/12C 14C BP  Cal. BP Cal. BC/AD
below MSL (%o)
BH I 6 437.5 Ly-9470 Marine shells 1.9 4931 62 5440e5040 3490e3090 BC
BHIX 8 167.5 Lyon-1798 (GrA 20857) Marine shells est 0 2370 50 2130e1860 180 BCe90 AD
BHIX 10 157.5 Lyon-1879 (Poz 998) Marine shells 2.26 2285 35 2000e1800 50 BCe150 AD
BHIX 24 86.5 Lyon-1878 (Poz 1016) Marine shells 0.35 2350 35 2090e1870 140 BCe80 AD
BHIX 36 236.5 Lyon-1796 (GrA 20859) Marine shells 1.61 2340 80 2180e1750 230 BCe200 AD
BHIX 35 273 Lyon-1797 (GrA 20858) Marine shells est 0 2240 60 1990e1690 40 BCe260 AD
BHIX 44 402.5 Lyon-1876 (Poz 1004) Marine shells 1.37 3640 50 3680e3400 1730e1450 BC
BHIX 47 477.5 Lyon-1877 (Poz 1002) Marine shells 1.58 4410 40 4720e4420 2770e2470 BC
BH VIII 6 350 Lyon-1799 (GrA 20809) Marine shells 1.88 4220 50 4450e4140 2500e2190 BC
BH VIII 10 472.5 Lyon-1728 (OxA) Marine shells 1.16 4060 40 4230e3960 2280e2010 BC
BH VIII 14 737.5 Lyon-1729 (OxA) Marine shells 3.57 5955 45 6470e6270 4520e4320 BC
BH VIII 16 780 Lyon-1730 (OxA) Marine shells 1.11 6030 45 6580e6310 4630e4360 BC
BH XV 3 161 Poz-13012 1 Venerupis rhomboides 1.9 2255 35 1970e1760 20 BCe190 AD
BH XV 12 247.5 Poz-13374 2 grape seeds 27.3 2515 30 2740e2480 790e530 BC
BH XV 17 315 Poz-13375 Charcoal 29.5 3385 35 3720e3550 1770e1600 BC
BH XV 24 367.5 Poz-13013 1 Nassarius reticulatus 4.4 3670 40 3690e3450 1740e1500 BC
BH XV 28 425 Poz-13006 2 Glycymeris glycymeris (juvs.), 8 4840 40 5280e5000 3330e3050 BC
1 Lucinella divaricata
BH XV 31 555 Poz-13007 5 Bittium reticulatum 3.6 5180 50 5650e5430 3700e3480 BC

pebbles are encrusted with marine fauna such as Serpulae. In The sedimentological data concur a middle energy beach
core BH I, this grades into a silty sand with a medium sorting environment. In core BH IX, the bottom of the unit is constrained
index of 0.88. to 4410  40 BP (2750e2480 cal. BC). Two Loripes lacteus
shells from unit C of BH I yielded a radiocarbon age of
5.1.2. Unit C2dPocket beach/proto-harbour 4931  62 BP (3475e3070 cal. BC). Molluscan taxa from
Unit C2 is characterised by a shelly sand unit, with poor sort- the following assemblages are attributed to this unit (Figs.
ing indices (1.17e1.27) and sand modal values of 250e200 mm. 5, 8 and 11): subtidal sands assemblage (Bulla striata and

Upper muddy-sand Upper

assemblage clean-sand Hard substrate
in sheltered areas assemblage SSHS Subtidal sands assemblage Lagoonal Mixt.
BH I assemblage Silty or
muddy-sand
+ 1.7 assemblage

+1
Infill Sediment
texture

Unit
Tests
Biological mean 10 20
0 sea level
Sands
Islamic
shards
A
-1
Muddy sands

-2
Depth (m)

Plastic clays B1

-3

Silty sands B2
and clays
-4
4931 ± 62 BP
3490 - 3090 BC

-5 C
Shelly sands

10 0 80 20 0 40 5 0 50 0 80 5 5 10 0 80 0 80 0 30 0 30 0 70 20 0 50 5 5 0 30

-6 Relative abundance (%) SSHS: Subtidal sands and hard

substrates

Silts and
sands Gravels Silts and clays Sands
-7 Transgressive
pebbles
Substratum

Fig. 5. Molluscan macrofauna from core BH I (northern harbour).

1520 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535

BHI
+ 1.7

Brackish lagoonal Marine lagoonal Coastal Marine

+1
Infill Sediment Ostracod

Unit
texture
groups
25 50 75
25 50 75 %
Biological mean
0 sea level
Islamic
shards

Muddy A
-1
Depth (m)

sands

-2
Plastic
clays
B1

-3

B2
Silty sands
-4
and clays

Shelly sands 0 90 0 90 0 30 0 30 10 10
Relative abundance (%)

Gravels

Sands Brackish lagoonal Coastal

Silts and clays Marine lagoonal Marine

Fig. 6. Ostracod microfauna from core BH I (northern harbour).

Mysia undata), upper clean-sand assemblage (Smaragdia vir-
idis, Nassarius pygmaeus, Neverita josephinia, and Chamela
gallina) and the upper muddy-sand assemblage in sheltered
areas (Loripes lacteus and Cerithium vulgatum). The ostracod
fauna comprises taxa from the marine lagoonal (Loxoconcha
sp. and Xestoleberis spp.) and coastal (Aurila spp., Urocy-
thereis sp. and Heterocythereis albomaculata) domains with
some marine taxa being drifted in (Figs. 6, 9 and 12). These
biostratigraphic data are analogous to a pocket beach, shel-
tered by the sandstone ridge.
Bela ginnania and Mitra cornicula), upper muddy-sand assem-
blage in sheltered areas (Loripinus fragilis, Nassarius corniculus
and Cerithium vulgatum), upper clean-sand assemblage (Nas-
sarius reticulatus) and the silty or muddy-sand assemblage
(Glycymeris glycymeris). The ostracod data evince continued
domination of marine lagoonal and coastal taxa, with a rise in
the brackish lagoonal taxon, Cyprideis torosa, towards the top
of the unit in BH IX. These all corroborate a semi-sheltered
environment that served as a proto-harbour during the Bronze
Age [30,62].

5.1.3. Unit C1dArtificial Bronze Age cove
A fall in energy dynamics is translated by a rise in the silts
fraction (up to 59%). Medium sands dominate, with a well to
medium sorted sediment. In BH IX, the base of the unit is con-
strained to 3640  50 BP (1730e1450 cal. BC), a date con-
firmed by data from BH XV (3670  40 BP or 1740e1500
cal. BC). We interpret this unit as corresponding to the Middle
Bronze Age to Late Bronze Age proto-harbour, with possible
reinforcement of the sandstone ridge improving the anchorage
quality. Small boats would have been hauled onto the beach
face, with larger vessels being anchored in the embayment.
Molluscan taxa include tests from the subtidal muds assem-
blage (Odostomia conoidea and Haminœa navicula), subtidal
sands assemblage (Tricolia pullus, Mitra ebenus, Rissoa dolium,
5.1.4. Unit B2dClosed Phoenician to Roman harbours
In unit B2, there is a net change in sedimentary conditions
with a marked shift to silts and fine grained sands. Modal grain
size values of 200e160 mm for the sand fraction and medium
sorting indices are concomitant with a low energy environ-
ment. Dominant molluscan groups include the lagoonal as-
semblage (Cerastoderma glaucum, Parvicardium exiguum
and Scrobicularia plana), the upper clean-sand assemblage
(Nassarius louisi and Nassarius pygmaeus) and the upper
muddy-sand assemblage in sheltered areas (Gastrana fragilis).
Rise in lagoonal taxa is in compliance with anthropogenic
sheltering of the environment by harbourworks. Ostracod
fauna is poor, with less than 50 tests per 10 g of sand, and
dominated by taxa from the brackish lagoonal (Cyprideis
 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535 1521

BH IX
+ 1.7

+1
Infill
Modal Sorting Index of Mode
Sediment Sand

Unit
Grain size grain in
texture texture index asymmetry
histograms (Weydert, 1973)

80 0
60 0
12 0

0
0
00
0
5
0
-0.4

-0.2

0
5 0 00 5 0
25 50 75 25 50 75

0.2
0.4
0.6

1.2
0.8

0.2
µm

53
40
0 1 5 06 2 0 0 3 0

16
0 0.2 0.4 0.6 -0.6 0.4

0
3 2 21 1 1 8 6 5

1
IX-1
Biological mean
0
0m sea level

%
µm
Muddy A
sands IX-29

2350 ± 35 BP %
-1 140 BC-80 AD
IX-6
Plastic
B1
%

clays
Depth (m)

2285 ± 35 BP µm
2370 ± 50 BP
180 BC-90 AD
-2
Marine IX-5
muds
%

B2
2340 ± 80 BP µm
230 BC-200 AD
2240 ± 60 BP
-3
40 BC-260 AD IX-33
%

µm
Sands C1
IX-42
%

3640 ± 50 BP
-4 1730-1450 BC µm

Shelly IX-47
sands C2
%

4410 ± 40 BP
µm

-5
2770-2470 BC
Transgressive D
pebbles

Sandstone Gravels Coarse sands [2mm;500µm]

substratum
Sands Medium sands [500µm;200µm]

Silts and clays Fine sands [200µm;50µm]

Fig. 7. Sedimentology of core BH IX (northern harbour).

torosa) and marine lagoonal (Xestoloberis sp. and Loxoconcha
sp.) ecological groups, all indicative of a protected environ-
ment. In BH XV, peaks of Aurila convexa are concomitant
with a proximal shoreface. Drifted-in valves of marine taxa
(Semicytherura sp., Microcytherura sp. and Hemicytherura
sp.) attest continued connection with the open sea and offshore
marine dynamics. In cores BH IX and BH XV, radiocarbon
dates constrain the chronology of the unit to 2515  30
(790e530 cal. BC) and 2340  80 BP (230 cal. BCe200
cal. AD), or the Phoenician/Persian to Roman periods. The
lower foundation courses of Zire Island’s jetties have also
been dated to the Persian period [11,31]. Persistent age-depth
anomalies concur analogous data in Tyre’s ancient harbour,
where strong chronostratigraphic evidence for dredging has
been detailed from the Roman period onwards.
5.1.5. Unit B1dClosed late Roman and early Byzantine
harbours
Accentuation of these low energy conditions is manifested
in unit B1 by a plastic clays unit. Throughout much of this fa-
cies, the silts and clays fraction comprise >90% of the aggre-
gate sediment. Radiocarbon ages are often stratigraphically
incoherent for this unit. Lagoonal (Cerastoderma glaucum
and Parvicardium exiguum) and upper muddy-sand assem-
blage in sheltered areas taxa (Cerithium vulgatum, Venerupis
rhomboides, Loripes lacteus and Macoma cumana) continue
1522 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535

Upper
muddy-sand
BH IX assemblage Silty or

Algae
LER
Upper clean-sand in sheltered Subtidal sands Hard substrate muddy-sand
+ 1.7 assemblage areas assemblage HP IM Mixt Lag MBA 1. assemblage assemblage No ecological significance

+1
Infill Sediment
texture
25 50 75

Unit
Biological mean Number of
0 sea level tests
20 60 100
Muddy
sands A

2350 ± 35 BP
-1 140 BC-80 AD
Plastic
Depth (m)

clays B1
2285 ± 35 BP
2370 ± 50 BP
180 BC-90 AD
-2 Marine
muds
2340 ± 80 BP B2
230 BC-200 AD

2240 ± 60 BP
40 BC-260 AD
-3
C1
Sands

3640 ± 50 BP
-4 1730-1450 BC

Shelly
sands C2
4410 ± 40 BP
2770-2470 BC
-5
Transgressive
pebbles Relative abundance (%) LER: Large ecological repartition
Sandstone Gravels IM: Subtidal muds
Lag: Lagoonal
substratum
Sands MBA: Mud bottom assemblage
Mixt: Mixiticole
Silts and clays 1. Subtidal sands and hard
subsrates

Fig. 8. Molluscan macrofauna from core BH IX (northern harbour).

BH IX
+ 1.7 Freshwater Brackish lagoonal Marine lagoonal Coastal Marine

+1 Infill
Ostracod

Unit
Sediment
texture
groups
Biological mean
0 sea level
25 50 75 %

Muddy
sands No ostracods A
2350 ± 35 BP
-1 140 BC-80 AD
Plastic
Depth (m)

clays B1
2285 ± 35 BP
2370 ± 50 BP
180 BC-90 AD
-2 Marine
muds
2340 ± 80 BP B2
230 BC-200 AD

2240 ± 60 BP
40 BC-260 AD
-3
Sands C1
No ostracods
3640 ± 50 BP
-4 1730-1450 BC

Shelly
sands C2
4410 ± 40 BP
2770-2470 BC
-5 0 90 0 30 10 0 50 0 90 10 0 40 20 10 5 5
Transgressive
pebbles Relative abundance (%)
Sandstone
substratum Gravels
Sands Brackish lagoonal Coastal
Silts and clays Marine lagoonal Marine

Fig. 9. Ostracod microfauna from core BH IX (northern harbour).

 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535 1523

BH XV
+1.7

Biological
0 mean
sea level

Infill

-1 Grain Sorting Mode

Sediment Sand Grain size Skewness

Unit
moyen in
index microns
texture texture histograms
25 50 75 25 50 75
BH XV 2

%
2255 ± 35 BP
20 BC - 190 AD
BH XV 5
Plastic

%
-2 clays B1
BH XV 9
%

2515 ± 30 BP
790 - 530 BC
Depth (m)

BH XV 14
%

Silty
sands
BH XV 17
B2
-3
%

3385 ± 35 BP
1770 - 1600 BC
BH XV 19
%

3670 ± 40 BP BH XV 24
1740 - 1500 BC
%

Grey BH XV 27 C1
sands
-4
%

4840 ± 40 BP BH XV 28
3330 - 3050 BC
%

BH XV 29
Shelly
%

sands C2
BH XV 30
-5
%

BH XV 31

5180 ± 50 BP
%

3700 - 3480 BC D
Clay
substratum
-6

Gravels Coarse sands [2mm;500µm]

Sands Medium sands [500µm;200µm]

Silts and clays Fine sands [200µm;50µm]

Fig. 10. Sedimentology of core BH XV (northern harbour).

to characterise the in situ taxa, attesting confined harbour con-
ditions. Quasi-dominance of Cyprideis torosa, with a rich fau-
nal density, indicates an extremely sheltered, lagoon-like
harbour during the late Roman and early Byzantine periods.
5.1.6. Unit AdExposed Islamic harbour
The base of unit A is marked by a sudden rise in the sands
fraction to the detriment of the silts and clays. The relative
abundance of the gravels and sands fractions gradually in-
creases up the unit. Ceramics constrain this facies to the Is-
lamic period. A shift from lower energy modal sand values at
the base (100 mm) to middle energy values (up to 315 mm)
at the top, could substantiate a gradual demise in harbour
maintenance. Biostratigraphic data affirm a reopening of the
environment to the influence of offshore marine dynamics,
with taxa from the subtidal sands assemblage, the hard sub-
strate assemblage and the upper muddy-sand assemblage.
The ostracod suites manifest an increase in marine lagoonal
(Loxoconcha sp. and Xestoleberis sp.) and coastal taxa (Aurila
spp.) to the detriment of Cyprideis torosa. We posit that this
could correspond to the gradual demise of Sidon as a commer-
cial centre during the Islamic period.
5.2. Southern cove facies
5.2.1. Unit CdHolocene transgression
The sandstone substratum is overlain by a thin pebbly sand
unit, dated 6030  45 BP (4630e4360 BC) and marking the
1524 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535

BH XV
+1.7
1. Silty or muddy-sand assemblage
2. Algae and hard substrate assemblage

Biological Upper muddy-sand

ae
Subtidal sands Subtidal sands Upper clean-sand Hard substrate

xt.
assemblage
0 mean

Alg
1 2

Mi
and hard substrates assemblage assemblage in sheltered areas assemblage Lagoonal
sea level

Infill

-1 Sediment Species

Unit
texture diversity
25 50 75 2 4 6 8 10 12

2255 ± 35 BP
20 BC - 190 AD
Plastic B1
-2 clays
2515 ± 30 BP
Depth (m)

790 - 530 BC

Silty
sands B2
-3
3385 ± 35 BP
1770 - 1600 BC

3670 ± 40 BP
1740 - 1500 BC
Grey C1
-4 sands
4840 ± 40 BP
3330 - 3050 BC

Shelly
sands C2
-5

D
5180 ± 50 BP
3700 - 3480 BC 7 14 21 28 36 42
Clay
substratum Taxa species
-6 Relative abundance (%)
diversity
Gravels
Number
Sands
Silts and clays

Fig. 11. Molluscan macrofauna from core BH XV (northern harbour).

Holocene marine transgression of the cove (Fig. 13). This sub-
sequently grades into a coarse shelly sand fraction, constrained
to 5955  45 BP (4520e4320 cal. BC) and typical of the sub-
tidal zone. Bittium spp., analogous of subtidal sands, character-
ises the molluscan fauna (Fig. 14). Loxoconcha spp. and
Xestoleberis aurantia dominate the ostracoda and concur
a shallow water marine embayment (Fig. 15).
5.2.3. Unit AdPrograding shoreline
Unit A comprises a shelly sands unit. An important rise in
the gravels fraction (21e40%), and the much coarser nature of
the sands is concomitant with a prograding shoreline. The
molluscan fauna is poor and essentially comprises badly pre-
served tests and shell debris, reworked by the action of the
swash. The higher energy dynamics of the swash zone were
not conducive to ostracod test preservation.

5.2.2. Unit BdBronze Age pocket beach 6. Discussion

The onset of unit B postdates 5955  45 BP, and comprises
a medium to fine-grained sand unit. Sample BH VIII 10
yielded a radiocarbon age of 4060  40 BP (2280e2010 cal.
BC). The molluscan fauna is diverse with tests from a range
of ecological contexts including subtidal sands, upper
muddy-sand assemblage in sheltered areas, upper clean-sand
assemblage, silty or muddy-sand assemblage and the lagoonal
assemblage. Marine lagoonal ostracod taxa persist into the
unit, and are accompanied by a gradual rise in coastal taxa
(Aurila spp., Urocythereis spp. and Cushmanidea elongata).
Sporadic tests of marine species indicate continued communi-
cation with the open sea.
6.1. Chronostratigraphy: ancient harbour technology
translated using sediment archives
Our geoarchaeological datasets elucidate a complex history
of coastal change and human occupation. Investigation of the
coastal archives has expounded understanding of Sidon’s mar-
itime history between the Bronze Age and Medieval periods.
Multidisciplinary investigations demonstrate that harbour
management advances are clearly translated by distinctive sed-
imentary facies and faunal suites. The sedimentary history of
the harbour details six periods.
 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535 1525

BH XV
+1.7

Biological Brackish lagoonal Marine lagoonal Coastal Marine Unkown

0 mean
sea level

Infill

Ostracod

Unit
-1 Sediment groups
texture
25 50 75 25 50 75 %

2255 ± 35 BP
20 BC - 190 AD
Plastic B1
-2 clays
2515 ± 30 BP
Depth (m)

790 - 530 BC

Silty No ostracods
sands B2
-3
3385 ± 35 BP
1770 - 1600 BC

3670 ± 40 BP
1740 - 1500 BC No ostracods C1
-4 Grey
sands
4840 ± 40 BP
3330 - 3050 BC

Shelly C2
-5 sands

5180 ± 50 BP
D
3700 - 3480 BC
0 90 0 40 0 50 10 0 90 10 10 0 40 5 5 5 0.1 1 10 100 1000
Clay Number of ostracods per 10 g
-6 Gravels Relative abundance (%)
substratum of sand (log scale)
Sands
Brackish lagoonal Coastal Unkown
Silts and clays
Marine lagoonal Marine

Fig. 12. Ostracod microfauna from core BH XV (northern harbour).

6.1.1. Bronze Age proto-harbour phase
At the time of Sidon’s foundation, during the third millen-
nium BC, maritime harbour technology was still very much in
its infancy [27,43]. Existing Bronze Age evidence from the
Levant shows a clear pattern of environmental determinism,
whereby coastal populations founded settlements in proximity
to naturally occurring anchorages such as leaky lagoons, estu-
aries and pocket beaches [60,61]. In Bronze Age Sidon, the
northern and southern pocket beaches, either side of the prom-
ontory, were ideally predisposed to serve as proto-harbours.
The northern cove, with its shelly medium sands facies, af-
forded the best natural shelter for larger merchant boats during
storms. Conversely, the wide sandy beaches prevalent in the
southern cove would have accommodated smaller vessels,
drawn from the water onto the beachface. This phenomenon
is still practiced today throughout the Mediterranean by fisher-
men with light, shallow draught vessels.
6.1.2. Dating the beginnings of the artificial harbour
Towards the end of the Late Bronze Age and the Early Iron
Age (w1200e1000 BC), expanding international trade
prompted coastal populations into modifying these natural an-
chorages [43]. Exempli gratia, the Phoenician mole at Atlit has
been constrained to the ninth century BC and is a paradigm of
artificial harbour infrastructure in the Levant [35]. Another
good archetype derives from Tabbat el-Hammam in Syria,
where a quasi-identical mole has been dated to ca. the ninth/
eighth centuries BC [8].
More speculative examples of harbour infrastructure are
known from the Levantine coast [30]. The earliest harbour
quays at Tell Dor have been attributed to the thirteenth century
BC [59,60,62]. Also in Israel, the Middle Bronze Age site of
Yavne-Yam attests the presence of submerged boulder piles,
used to improve the quality of the ancient anchorage (Marcus,
personal communication). Although open to fervent debate, texts
from Ras-Shamra and Ugarit could comply a fifteenth century
BC age for the foundation of Minet el-Beida on the Syrian
coast [66,70].
In Sidon’s northern harbour, transition from shelly to fine-
grained sands appears to be the earliest granulometric mani-
festation of human coastal modification. A single radiocarbon
date from core BH XV constrains this alteration to the Middle
Bronze Age (w1700 cal. BC) and must be confirmed by fur-
ther data. The Phoenicians cleverly quarried sandstone ridges
to form artificial quays. Surplus blocks were frequently reem-
ployed to either construct (i.e. Arwad) or reinforce (i.e. Sidon
and Tripoli) a seawall [11,12,73e75] (Fig. 16). In Sidon’s
northern harbour, modernisation means that these vestiges
1526 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535

BH VIII
+3

0 Biological
mean
sea level Infill
Grain Sorting Mode
Sediment Skewness in

Unit
Sand Grain size moyen index
texture microns
texture histograms

+0.2

+0.4
0.4

-0.4

-0.2
0.2

0.6

0.4

0.8

1.2

1.6
-1 25 50 75 25 50 75

10 0
80 0
12 0
5 0 0 0 5 0 0 3 0

0
0
0
0
5
0
0
31 25 20 16 12 10 8 6 5 µm

63
50
40
16
BH VIII 1

%
BH VIII 3
Shelly
-2
A

%
sands
BH VIII 5
%

-3
BH VIII 6
Depth (m)

4220 ± 50 BP
%

2500 - 2190 BC

BH VIII 7
-4
%

4060 ± 40 BP BH VIII 10
2280 - 2010 BC
%

-5
B
Grey
sands
BH VIII 12
-6
%

BH VIII 13
%

-7
5955 ± 45 BP BH VIII 14 BH VIII 15
4520 - 4320 BC
%

Shelly sands C
6030 ± 45 BP
-8
Trangressive
Substratum
(Ramleh)
pebbles
Gravels [2mm;500µm] Coarse sands

Sands Medium sands [500µm;200µm]

Silts and clays Fine sands [200µm;50µm]

Fig. 13. Sedimentology of core BH VIII (southern cove).

are no longer visible, although they have been described by concurrent with two complimentary dynamics: (1) modest
d’Arvieux [2], Renan [64], Lortet [42] and surveyed by Poide- artificial harbourworks during the late Middle Bronze Age
bard and Lauffray [57]. and Late Bronze Age; and/or (2) intense dredging during the
The lee of Zire was also exploited as a deep-water Roman and Byzantine occupations (see below).
anchorage, or outer harbour, at this time although the island’s
two jetties date from the Persian period. In fair to medium 6.1.3. Absence of Iron Age sedimentary archives
weather, large merchant vessels would have loaded and un- Given the relative absence of pre-Hellenistic facies, high-
loaded goods at this geomorphologically predisposed hub, resolution reconstruction of the northern harbour’s Phoenician
their cargos ferried to and from the shoreline by lighters history is problematic. Advanced harbour management tech-
[43,76]. A whole suite of harbourworks, including seawalls, niques during the Roman and Byzantine periods culminated
quays and mooring bits have been dug into the Quaternary in the repeated dredging of Sidon’s northern harbour, re-
sandstone, rendering Zire an integral component of Sidon’s moving this strata from the geological record (Fig. 17). In
port system. cores BH I and BH IX, there is a clear sediment hiatus be-
The bio-sedimentological datasets show, however, that it tween w1700 and 1500 cal. BC and the Roman period. As
was the northern harbour, naturally protected from the open in Tyre, reworking of the marine bottom is delineated by the
sea by a sandstone ridge, which became the city’s major chronostratigraphic dataset and supports unequivocal evidence
port basin. Conversely, there is no evidence for harbourworks from Marseilles [36,37] and Naples [32,33]. In BH XV, a more
in the southern cove. Difficulty in dating the first phase of ar- coherent and continuous chronology suggests that this area
tificial confinement, in both Sidon and Tyre, appears was less affected by dredging activity. The data evoke
 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535 1527

Subtidal sands Upper muddy-sand Silty or muddy-sand

BH VIII and Subtidal sands assemblage Upper clean-sand Hard substrate assemblage N ecological
hard substrates assemblage in sheltered areas assemblage assemblage 1 2 3 4 significance
+3

ilis

um
llin cf. ep abi rag
lli

Tellina jos vari us) f m

0 Biological

rei

Te erita ntes ripin neu

tat
uic ba
a n se hin lis

r is
m
meansea

s
)

ta
lat

My dor rac ryp eola s

Gi ama muiterraosus

ta
ibb nts

Co ilidaegraeetti oide

os

is
s uu

Di bul cf. ltilin neu

eq gib

me
a

ta
lac m e lata
Lo rvic la au atum

itid rra ia

Di us v trian ata
Na ax s anat atus

a
Ra lodo ruco ular
Ne hidoia (L iterr
um
level

ca
teu xig

Cy nus us g me

Ri ssar ma s tund
en lin vir a

a ta
Alv a de eolaidis

Lu obia glycy
Ce hidaeticu ta

ul
atu

ina )
k h

Na hito ta ro sa
Ct soa dia triat

ic
Tr ania cuss ta

la t

dis

un
lat

ca

g
Go nella lvae
a

tri
rip ard ric

g
a

.
Pa icu vulg

Ve alia ivar
o

p
Br odon me

Co sari .(fra
l
icu

btr
icu

Do lina mis

liu loi
Sediment

o a g

Co id ae
Ch ar ed
es iu

s
u

ae
nta (A

d
Ri ara cf.

su
ret

r et

eri
e

nid ae
Infill

m
Ri hium

p
e

ss ii d
texture

n pl

n
r

ac t

ae

iid
Te x s
An tidiu
g

t a
De rbula
Smllaria

ym
Species

la
m

Unit
rith

isu
o
od
-1

na
tiu

tiu

rit

dr
ng

yc
oc
25 50 75

ci

p
v

n
diversity

p
t
s

Ce
Do
Bu

Hy

Sp
Bit

Bit

Le

Gl
4 8 12 16

-2 Shelly
A
sands

-3
Depth (m)

4220 ± 50 BP
2500 - 2190 BC
-4

4060 ± 40 BP
2280 - 2010 BC
-5 B

Grey
-6 sands

-7
5955 ± 45 BP
C
4520 - 4320 BC
20 40 60 80
6030 ± 45 BP 20 20 5 10 0 30 0 80 0 30 0 30
-8 Taxa species
Trangressive Relative abundance (%) diversity
Substratum
(Ramleh)
pebbles 1. Lagoonal Number
2. Mixticole
Gravels Silts and clays Sands 3. Algae
4. Mud bottom assemblage

Fig. 14. Molluscan macrofauna from core BH VIII (southern cove).

advanced harbourworks during the Phoenician and Persian
periods.
Siltation, notably under deltaic and urban contexts, was
a well-recognised problem in antiquity with four sedimentary
sources of note: (1) local watercourses; (2) regional longdrift
currents; (3) erosion of adobe constructions and urban runoff;
and (4) use of the basin as a base-level waste dump. Sidon’s
gravels fraction from the Roman period comprises a whole
suite of discarded objects, trapped at the bottom of the basin,
including ceramics, wood, seeds, leather artefacts etc. Indeed,
an inscription from Roman Ephesus, demanding citizens not to
throw waste into the port, attests that ancient societies must
have been acutely aware of this problem (Die Inschriften
von Ephesos, Bonn, I, 1979, no. 23).
It is postulated that extensive dredging during the Roman
and Byzantine period explains (1) the observed stratigraphic
hiatus and (2) dating inversions. Previously, these problems,
in the absence of robust chronological frameworks, were
most often ignored or left unexplained.
6.1.4. The Roman revolution
By Roman times, the discovery and use of hydraulic con-
crete greatly enhanced engineering possibilities (cf. Vitruve,
V, 12), locally deforming coastal landscapes [9,51]. In effect,
at this time we observe pronounced transition from environ-
mental to anthropogenic determinism. The Romans were
able to conceive long breakwaters or offshore harbour basins,
Caesarea Maritima being an example par excellence [5]. All-
weather basins could be constructed at locations where no nat-
ural roadstead existed. During this technological revolution,
Sidon’s northern harbour underwent significant changes with
the edifice of an inner artificial mole perpendicular to the sand-
stone ridge [57]. This yielded an extremely well-protected ba-
sin, translated in the geological archive by a silt facies
containing lagoonal molluscs and microfossils.
Under these closed conditions sedimentation rates rose
significantly (w1 mm/year during the mid-Holocene com-
pared to w10 mm/year for the Roman period), not least
because of the overriding confinement, but also linked to
increasing human use and abuse of the surrounding watershed
that flushed sediment into coastal depocentres. This dual phe-
nomenon, increased confinement coupled with a rise in sedi-
ment yields, is consistently observed in base-level harbour
basins throughout the circum Mediterranean. In its most acute
form the ensuing coastal progradation led to harbour land-
locking, isolating basins many kilometres from the sea.
1528 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535

BH VIII Freshwater Marine lagoonal Coastal Marine

+3

nia
ta
ata

ci a

vo
a

ng
Shelly

.
n ti

sp
pa
Sediment

the p. fas
ea p .
el o
ura

g
.
pp
sands

is
p. on
texture

tta
nid sp

cy a s is

e re

e re
a
sa
as

.
sp ris l

ere
ru r
mi ett de

sp
ma eis

.
25 50 75 Ostracod

yth

Lo td cyth
Ju eb spp
eri
ch

i
Se her er

la
rila be

yth
sh her
-3

eb

Unit
n

oc

re l
h
groups

Austole
co

Pcgoso
Tr irdia

Faculic
t
Cuocyt

Cy ocy
l

Px llist

the
sto
xo

Cydt
o

Bat

Ncl
Ca
Ne
Xe
Xe

i
Lo

Ur
Ily
25 50 75 %
4220 ± 50 BP
2500 - 2190 BC

-4
Depth (m)

4060 ± 40 BP
2280 - 2010 BC
-5
B
Grey
sands

-6

-7

5955 ± 45 BP
4520 - 4320 BC C
Shelly sands
6030 ± 45 BP
0 30 0 40 0 90 10 5 5 0.1 1 10 100 1000
-8
Relative abundance (%)
Substratum Number of ostracods per 10 g
(Ramleh) (log scale)
Trangressive
pebbles
Gravels Silts and clays Freshwater Coastal
Sands Marine lagoonal Marine

Fig. 15. Ostracod microfauna from core BH VIII (southern cove).

Celebrated examples are known from the Aegean Anatolian Sidon is marked by advanced reinforcement of the antecedent
deltas and include the sites of Troy, Ephesus, Priene and port infrastructure, with a notable persistence of Roman tech-
Miletos [10,39]. nology and its opulent legacy of engineering achievements
In addition to artificial dredging, engineering solutions to [38]. This is corroborated by a plastic clay unit with diagnostic
the siltation problem have been asserted, although many of lagoonal macro- and microfossils, typical of pronounced con-
these remain speculative [4]. At Sidon, Poidebard and Lauf- finement. These geological data support archaeological evi-
fray [57] identified a flushing channel carved into the sand- dence from Beirut’s Byzantine harbour suggesting that the
stone and linking the northern basin with the open sea. Levantine coast was still an important trade zone at this
Undated, the two scholars hypothesised this to be an ad hoc time [68]. Such a trade apex, coeval with advanced port infra-
desilting channel, dug to generate current through the inner structure and management techniques, leads us to propose
harbour and alleviate the effects of sediment deposition. The a harbour apogee for Sidon during the Byzantine period.
stratigraphy shows dredging and coeval desilting infrastructure
to have been insufficient in completely eradicating the prob-
lem; two thousand years later, rapid silting up means that 6.1.6. Harbour semi-abandonment phase
the majority of the ancient basin is now buried beneath the Radical coastal changes are witnessed during the Islamic
Modern city centre. period, with a gradual reopening of Sidon and Tyre’s northern
harbours. We advance three hypotheses to explain these data,
namely (1) cultural, (2) tectonic and (3) tsunamogenic.
6.1.5. Byzantine apogee and persistence of Roman (1) Historians traditionally argue that the Byzantine crisis
technology (sixth century AD) and ensuing seventh century AD Islamic
Whilst Bronze Age populations benefited from Sidon’s conquest engendered profound changes in the eastern Mediter-
geological endowment, Byzantine societies inherited the ranean’s trade network [53] (Fig. 18). During and after these
Romans’ rich maritime savoir faire. The Byzantine period in permutations, it is speculated that harbour infrastructure fell
 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535 1529

Fig. 16. Sidon’s northern harbour and its ancient seawall. Before recent construction work an uplifted marine notch was observed at w50 cm above MSL. All
photos C. Morhange (1998).

into a state of disrepair. A priori our sedimentological data are
broadly consistent with these historical interpretations. Never-
theless, recent archaeological and historical research tends to
moderate the premise of a general decline of ‘Syrian’ harbours
at this time [6,28,41,72]. Historians believe the Levantine
coast to have been the cradle of Islamic maritime develop-
ment, control which opened the gates of the Mediterranean.
For many Medieval authors, Tyre epitomises the harbour
model par excellence [7]. Historical sources evoke three piv-
otal Levantine harbours of note, Acre, Tyre and Tripoli,
whereas the other port sites seemingly disappear from the mar-
itime map. Acre and Tyre were, for instance, important centres
of naval construction during the ninth to tenth centuries AD.
Our data indicate rapid coastal progradation from the sixth
century AD onwards, entraining the deformation and
dislocation of Sidon’s harbour. Although smaller, it is difficult
to accurately constrain the exact dimensions of this port as the
post-Byzantine sediment record lies beneath the present basin
and is not accessible at present.
(2) At Tyre, port opening may have been amplified by the
w3 m late Roman tectonic collapse of the Phoenician island
(see Fig. 17). Its northern Roman (?) mole is currently 2.5 m
below present sea level, translating a subsidence of w3e
3.5 m. On the southern shore, walls and drowned quarries at
2.5 m below MSL have also been discovered, and similar
subsidence is translated in the city’s coastal stratigraphy. By
contrast, sea-level data from Sidon shows that the magnitude
of crustal mobility is inferior to Tyre, w50 cm since antiquity,
yet the same coarse sand facies is persistently observed. These
data chronologically contradict, at least locally, the Early Byz-
antine Tectonic Paroxysm hypothesis dated to the fourth to
sixth centuries AD [54e56]. In effect, the opening of Tyre
1530 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535

Calibrated radiocarbon years

9000 8000 7000 6000 5000 4000 3000 2000 1000 0
0

200 200

400 400
D ON
SI
Depth below biological
mean sea level (cm)

600 600

RE
TY
800 800

1000 1000

1200 1200
7000 6000 5000 4000 3000 2000 1000 BC/AD 1000 2000
Calibrated radiocarbon years B.C./A.D.

Natural cove/protoharbor phase

Roman and Byzantine protected and dredged harbor
‘Global’ Sea-level envelope
Sidon dates and 95% confidence levels

Tyre dates and 95% confidence levels

~3 m post-Roman tectonic collapse (Tyre)

Fig. 17. Chronostratigraphic evidence for Roman and Byzantine dredging of Sidon and Tyre’s ancient harbours. The older radiocarbon group corresponds to a nat-
urally aggrading marine bottom. Quasi-absence of a chronostratigraphic record between BC 4000 to 500, coupled with persistent age depth inversions, are inter-
preted as evidence of harbour dredging (adapted from [46]).
 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535 1531

The Byzantine Empire in AD 527 The Islamic expansion AD 622 - 750

Byzantine rule

Islamic rule

BLACK BLACK
SEA SEA

Constantinople Constantinople

Sidon Sidon
MEDITERANNEAN Tyre MEDITERANNEAN Tyre

SEA SEA
N N
0 500 km 0 500 km

Fig. 18. Isolation of Sidon and Tyre from the Constantinople metropolis during the 7th to 8th centuries AD.

and Sidon appears to be later, after the sixth century AD.
Fig. 19 plots earthquake and tsunami events on the Levantine
coast illustrating that the fourth to eleventh centuries were
characterised by repeated seismic shocks, possibly provoking
partial harbour damage. According to data from various sour-
ces [52,65] it is interesting to note that during the EBTP a clus-
ter of five earthquakes 8 are documented on the Levantine
coast against a mere two during the period AD 600 to AD
1100 (scale sensu Plassard and Kogoj [52]).
(3) This discrepancy leads us to moderate tectonic collapse
in favour of documented tsunamogenic impacts. In effect,
more than ten tsunami struck the Levantine coast between
the fourth to eleventh centuries AD, severely damaging har-
bour infrastructure [1,34,71]. For the AD 551 event, John of

Year (AD)
300 400 500 600 700 800 900 1000 1100

12

10
(after Plassard and Kogoj, 1981)
Earthquake intensity

2
EARLY BYZANTINE TECTONIC
PAROXYSM RELATIVE DEMISE OF SIDON AND TYRE
0
300 400 500 600 700 800 900 1000 1100
Year (AD)

Earthquake (intensity > 8) Earthquake (intensity < 8) Tsunami

Fig. 19. AD 300 to 1000 earthquake and tsunami events affecting the Phoenician coast (data compiled from [1,34,52,65,71]).
1532 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535
Ephesus records that ‘‘[.] the sea withdrew and retreated
from the coastal cities of Phoenicia for a distance of nearly
two miles [.] A tremendous surge of the sea rushed up to re-
turn to its original depth’’ [34]. Notwithstanding these catas-
trophist scenarios Sidon and Tyre’s ports are still in use
today, 5000 years after their foundation.
6.2. Comparison between Sidon and Tyre
Analogous sedimentary stories are observed at both Sidon
and Tyre, reinforcing the premise that geoscience is a robust
archaeological tool. Nevertheless, Sidon’s northern harbour fa-
cies consistently comprises fine-grained material. During the
Roman period for example, Sidon’s harbour constitutes
>80% silts and fine sands whereas in Tyre the sand fraction
still predominates. How should such a sharp disparity be con-
strued? We advance two hypotheses, technological and geo-
morphological, both of which are not mutually exclusive: (1)
Tyre’s northern harbour is comparatively open, with a wide
w100e150 m channel entrance separating the two ancient
moles. This created increased exposure to outer marine dy-
namics. In contrast, the northern harbour of Sidon is quasi-
landlocked by three main obstacles, scilicet the sandstone
breakwater, the sea castle island and the inner mole. (2) Geo-
graphically, Tyre lies 9 km from the mouth of the Litani, Phoe-
nicia’s most important fluvial system. This river delivers
coarse sediment inputs to the coastal zone, constituting mainly
sands and gravels, trapped in base-level depocentres such as
harbour basins (Fig. 20). In contrast, Sidon is situated south
of the Awali river, a much smaller watershed yielding mainly
medium sands and silts.
Multiple similarities are recorded between these two sister
harbours. (1) Foremost, historical coastal progradation has led
to the isolation of the ancient harbours beneath the modern
city centres (Fig. 21). Around half of the basins are now
2002/12
N 13/7/9
0 1000 m
2002/18
= 18/7/1
W E
S
1-15
>15
knots
knots
Silts & clays Gravels
Northern harbour/
19/7/2
Sands
Fine sands Coarse sands
Medium sands
Fig. 20. Sedimentology of Litani sediments, north of the Tyrian peninsula (Landsat image).
buried beneath thick tracts of marine sediments, and the
historical coastline presently localised 100e150 m inland.
Such progradation has fossilised Sidon and Tyre’s archaeolog-
ical heritage offering significant scope for future research
and our understanding of Phoenicia’s maritime history [44].
(2) The difficulty in dating the first phase of artificial harbour
confinement translates modest harbourworks during the
Middle Bronze Age and the Late Bronze Age. Early seafarers
used semi-protected pocket beaches with little need to modify
the natural coastline. The weight of natural factors in influenc-
ing human exploitation of the coastal environment prevailed
until the Phoenician period. Only sediment sections yield
the resolution necessary to observe fine environmental details,
although these are logistically and financially unrealisable at
present.
7. Conclusion
In conclusion, we would like to insist on three research
advances.
(1) It has been demonstrated that harbour history can
be clearly chronicled by diagnostic litho- and biostratigra-
phies. Indeed, explicit use of the coastal geological record
has the possibility to greatly enhance our understanding of
human-environment interactions and their multiple facets;
we therefore postulate geoscience to be a powerful tool in
expounding the spatial organisation of harbour areas and their
coastal evolution through time [47]. Ancient harbours are also
appropriate for the analysis of numerous archaeological prob-
lems and cultural processes, providing a diversity of research
possibilities.
(2) For Sidon and Tyre, the biological proxies asseverate
a broadly similar chronostratigraphic pattern, under disparate
geomorphological and geodynamic contexts, a paradox
supporting the initial hypothesis of technologically forced
13/7/8
2002/15
2002/13
2002/11 & 11B
13/7/6
Litani
2002/10 2002/16
2002/14
N
 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535 1533

N Phoenician

m
5
seawall

m
N

0
T5
T1 T4
T2
T9 T6 0 200 m
eway
Tyre Ancient caus
island
Tombolo
Bronze Medieval
Age
castle
10 m

Submerged coastline
urban quarters
Buried
ancient

5m

0m
basin
0 1000 m

BH II
Ancient harbour's
maximum extension BH IV

Inner
northern
harbour
BH IX BH V
BH I BH VI
BH XV
BH III BH VII

BH XII
Sandstone
ridge BH XIV
BH XIII
BH X
Murex Tell

Open southern
Dominant harbour
swell BH XI

BH VIII

Sandstone
ridge

Dakerman

Fig. 21. Sidon’s reconstructed harbour limits in antiquity. Inset: Tyre’s northern harbour limits.

sedimentary changes. A clear transition is manifest in the stra- (3) Emerging geological data delineates a largely analogous
tigraphy from environmental determinism during the Bronze chronostratigraphic pattern for many of the Mediterranean’s
Age, through various intermediary phases, to full anthropo- ancient ports. Locally, the most important factors in explaining
genic determinism by the Roman period. coastal progradation are not small-scale relative sea-level
1534 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535
variations but anthropogenic sediment supply and harbour
management.
Acknowledgements
We thank the Directorate General of Antiquities (F. Husseini),
CEDRE (F60/L58), UNESCO World Heritage (CPM
700.893.1) and the Lebanese British Friends of the National
Museum for technical and financial support. N. M. benefited
from a Leverhulme Study Abroad Studentship. The authors
also wish to thank A. Borrut, M. Bourcier, N. Carayon, K.
Espic, J.-P. Goiran, N. Kashtan, J. Oleson and D. Sivan for
kindly reviewing earlier versions of the manuscript.
References
[1] N.N. Ambraseys, Data for the investigation of seismic sea-waves in the
Eastern Mediterranean, Bulletin of the Seismological Society of America
52 (1962) 895e913.
[2] L. d’Arvieux, Mémoires du chevalier d’Arvieux, envoyé extraordinaire
du Roi à la Porte, consul d’Alep, d’Alger, de Tripoli, et autres échelles
du Levant, contenant ses voyages à Constantinople, dans l’Asie, la Syrie,
la Palestine, l’Egypte et la Barbarie, la description de ces pays, les reli-
gions, les moeurs, les coutumes, le négoce de ces peuples, et leurs gou-
vernements, Jean-Baptiste Delespine, Paris, 1735.
[3] D.J. Blackman, Ancient harbours in the Mediterranean, part 1, The Inter-
national Journal of Nautical Archaeology and Underwater Exploration 11
(1982) 79e104.
[4] D.J. Blackman, Ancient harbours in the Mediterranean, part 2, The Inter-
national Journal of Nautical Archaeology and Underwater Exploration 11
(1982) 185e211.
[5] D.J. Blackman, in: A. Raban, K.G. Hollum (Eds.), Caesarea Maritima: A
Retrospective after Two Millennia, Brill, Leiden, New York, Köln, 1996.
[6] A. Borrut, L’espace maritime syrien au cours des premiers siècles de
l’Islam (VIIe-Xe siècle): le cas de la région entre Acre et Tripoli, Tem-
pora, Annales d’histoire et d’archéologie 10e11 (1999e2000) 1e33.
[7] A. Borrut, Architecture des espaces portuaires et réseaux défensifs du lit-
toral syro-palestinien dans les sources arabes (VIIe-Xie s.), Archéologie
Islamique 11 (2001) 21e46.
[8] R.J. Braidwood, Report on two sondages on the coast of Syria, south of
Tartous, Syria 21 (1940) 183e221.
[9] C. Brandon, in: A. Raban, K.G. Hollum (Eds.), Caesarea Maritima: A
Retrospective after Two Millennia, Brill, Leiden, New York, Köln, 1996.
[10] H. Brückner, M. Müllenhoff, M. Handl, K. van der Borg, Holocene land-
scape evolution of the Büyük Menderes alluvial plain in the environs of
Myous and Priene (Western Anatolia, Turkey), Zeitschrift für Geomor-
phologie N.F., 127 (Suppl.-Bd) (2002) 47e65.
[11] N. Carayon, L’ı̂le de Ziré à Sa€ıda: nouvelles données archéologiques,
Archaeology and History in Lebanon 18 (2003) 95e114.
[12] N. Carayon, J. Viret, in: F. Denise, L. Nordiguian (Eds.), Une aventure
archéologique: Antoine Poidebard, photographe et aviateur, Marseille,
Arles, Beyrouth, 2004.
[13] G. Contenau, Mission archéologique à Sidon (1914), Syria I (1920) 16e55
and 108e154.
[14] G. Contenau, Deuxième mission archéologique à Sidon (1920), Syria IV
(1924) 261e281.
[15] G. Contenau, Deuxième mission archéologique à Sidon (1920), Syria V
(1924) 9e23 and 123e134.
[16] J. Curtis, New excavations at Sidon, National Museum News 10 (2000)
27e40.
[17] C. Doumet-Serhal, SidondBritish Museum Excavations 1998e2003,
Archaeology and History in Lebanon 18 (2003).
[18] L. Dubertret, Carte géologique du Liban au 1/200,000, avec notice expli-
cative, Ministère des Travaux Publics, Beyrouth (1955).
[19] L. Dubertret, Introduction à la Carte Geologique au 1:50 000 du Liban,
Notes Mémoire Moyen-Orient 13 (1975) 345e403.
[20] M. Dunand, Chronique, Bulletin du Musée de Beyrouth III (1939)
79e81.
[21] M. Dunand, Chronique, Bulletin du Musée de Beyrouth IV (1940) 118.
[22] M. Dunand, Chronique, Bulletin du Musée de Beyrouth V (1941) 88e89.
[23] M. Dunand, Chronique, Bulletin du Musée de Beyrouth VI (1942e43)
82e83.
[24] M. Dunand, Rapport préliminaire sur les fouilles de Sidon en 1964e
1965, Bulletin du Musée de Beyrouth XX (1967) 27e44.
[25] M. El Amouri, M. El Helou, M. Marquet, I. Nourredine, M. Seco Al-
varez, Mission d’expertise archéologique du port sud de Tyr, sud Liban:
résultats préliminaires, Bulletin d’Archéologie et d’Architecture Libana-
ises HSII (2005) 91e110.
[26] K. Espic, C. Morhange, M. Bourcier, C. Bruzzi, P. Carbonel,
T. Nammour, C. Doumet-Serhal, Les ports antiques de Sidon: nouvelles
données paléo-environnementales, Archaeology and History in Lebanon
15 (2002) 28e36.
[27] D. Fabre, Seafaring in ancient Egypt, Periplus, London, 2004/2005.
[28] A.M. Fahmy, Muslim naval organisation in the eastern Mediterranean
from the seventh to the tenth century AD, National Publication and Print
House, Cairo, 1966.
[29] H. Frost, The Offshore Island Harbour at Sidon and Other Phoenician
Sites in the Light of New Dating Evidence, The International Journal
of Nautical Archaeology and Underwater Exploration 2 (1973) 75e94.
[30] H. Frost, in: V. Karageorghis, D. Michaelides (Eds.), Proceedings of the
International Symposium ‘Cyprus and the Sea’, University of Cyprus,
Nicosia, 1995.
[31] H. Frost, Installations on the ancient offshore anchorage at Sidon (the
rock island of Zire), National Museum News 10 (2000) 69e73.
[32] D. Giampaola, V. Carsana, in: E. Lo Sarde (Ed.), Eureka! il genio degli
antichi, Napoli, 2005.
[33] D. Giampoala, V. Carsana, G. Boetto, Il mare torna a bagnare Neapolis.
Parte II: dalla scoperta del porto al recupero dei relitti, L’archeologo sub-
acqueo 10 (3) (2004) 15e19.
[34] E. Guidoboni, A. Comastri, G. Traina, Catalogue of Ancient Earthquakes
in the Mediterranean Area up to the 10th Century, Istituto Nazionale di
Geofisica, Bologna, 1994.
[35] A. Haggai, Excavation of the Phoenician harbour at Atlit: findings and
reassessment of its foundation and purpose, Journal of the Tel Aviv Insti-
tute of Archaeology 33 (2006).
[36] A. Hesnard, in: L. De Maria, R. Turchetti (Eds.), Evolución paleoam-
biental de los puertos y fondeaderos antiguos en el Mediterráneo occi-
dental, Rubbettino Editore, Soveria Mannelli, 2004.
[37] A. Hesnard, in: A. Gallina Zevi, R. Turchetti (Eds.), Le strutture dei porte
e degli approdi antichi, Rubbettino Editore, Soveria Mannelli, 2004.
[38] R.L. Hohlfelder, Building harbours in the early Byzantine era: the persis-
tence of Roman technology, Byzantinische Forschungen Internationale
Zeitschrift für Byzantinistik 24 (1997) 367e389.
[39] J.C. Kraft, G.R. Rapp, I. Kayan, J.V. Luce, Harbor areas at ancient Troy:
Sedimentology and geomorphology complement Homer’s Iliad, Geology
31 (2003) 163e166.
[40] J. Laborel, F. Laborel-Deguen, Biological indicators of relative sea-level
variations and co-seismic displacements in the Mediterranean region,
Journal of Coastal Research 10 (1994) 395e415.
[41] A.R. Lewis, Naval power and trade in the Mediterranean AD 500-1100,
Princeton University Press, Princeton New Jersey, 1951.
[42] L. Lortet, La Syrie d’aujourd’hui, Voyages dans la Phénicie, le Liban et
la Judée, 1875e1880, Hachette, Paris, 1884.
[43] E. Marcus, in: E.C.M. van der Brink, T.F. Levy (Eds.), Egypt and the Le-
vant, Interrelations from the 4th through the Early 3rd Millennium BCE,
Continuum International Publishing Group, London, 2002.
[44] N. Marriner, C. Morhange, Under the city centre, the ancient harbour.
Tyre and Sidon: heritages to preserve, Journal of Cultural Heritage 6
(2005) 183e189.
[45] N. Marriner, C. Morhange, M. Boudagher-Fadel, M. Bourcier,
P. Carbonel, Geoarchaeology of Tyre’s ancient northern harbour, Phoeni-
cia, Journal of Archaeological Science 32 (2005) 1302e1327.
 N. Marriner et al. / Journal of Archaeological Science 33 (2006) 1514e1535
[46] N. Marriner, C. Morhange, Geoarchaeological evidence for dredging
in Tyre’s ancient harbour, Levant, Quaternary Research 65 (2006)
164e171.
[47] N. Marriner, C. Morhange, C. Doumet-Serhal, P. Carbonel, Geoscience
rediscovers Phoenicia’s buried harbors, Geology 34 (2006) 1e4.
[48] C. Morhange, J.-P. Goiran, M. Bourcier, P. Carbonel, J. Le Campion,
J.-M. Rouchy, M. Yon, Recent Holocene paleo-environmental evolution
and coastline changes of Kition, Larnaca, Cyprus, Mediterranean Sea,
Marine Geology 26 (2000) 205e230.
[49] C. Morhange, K. Espic, C. Doumet-Serhal, M. Bourcier, P. Carbonel, The
ancient harbours of Sidon, attempt at a synthesis (1998e2002), Archae-
ology and History in Lebanon 18 (2003) 71e81.
[50] C. Morhange, F. Blanc, M. Bourcier, P. Carbonel, A. Prone, S. Schmitt-
Mercury, D. Vivent, A. Hesnard, Bio-sedimentology of the late Holocene
deposits of the ancient harbor of Marseilles (Southern France, Mediterra-
nean sea), The Holocene 13 (2003) 593e604.
[51] J.P. Oleson, C. Brandon, S.M. Cramer, R. Cucitore, E. Gotti, R.L. Hohlfelder,
The ROMACONS Project: a Contribution to the Historical and Engineering
Analysis of Hydraulic Concrete in Roman Maritime Structures, International
Journal of Nautical Archaeology 33 (2004) 199e229.
[52] J. Plassard, B. Kogoj, Catalogue des seismes ressentis au Liban, Ann.
Sism. Observ. Ksara, CNRS, Beyrouth, 1981.
[53] H. Pirenne, Mahomet et Charlemagne, Les Presses Universitaires de
France, Paris, 1937.
[54] P.A. Pirazzoli, The Early Byzantine Tectonic Paroxysm, Zeitschrift für
Geomorphologie 62 (1986) 31e49.
[55] P.A. Pirazzoli, A review of possible eustatic, isostatic and tectonic
contributions in eight late-Holocene relative sea-level histories from
the Mediterranean area, Quaternary Science Reviews 24 (2005) 1989e
2001.
[56] P.A. Pirazzoli, J. Laborel, S.C. Stiros, Earthquake clustering in the East-
ern Mediterranean during historical times, Journal of Geophysical
Research 101 (B3) (1996) 6083e6097.
[57] A. Poidebard, J. Lauffray, Sidon, aménagements antiques du port de
Sa€ıda. Etude aérienne au sol et sous-marine, Imprimerie Catholique,
Beyrouth (1951).
[58] A. Raban, in: A. Raban (Ed.), Harbour Archaeology. Proceedings of the
first international workshop on ancient Mediterranean harbours, BAR
International Series 257, Caesarea Maritima, Haifa, 1985.
[59] A. Raban, Alternated river courses during the Bronze Age along the
Israeli coastline, Déplacements des lignes de rivage en Méditerranée
d’après les données de l’archéologie, CNRS, Paris, 1987.
1535
[60] A. Raban, The Harbor of the Sea People at Dor, Biblical Archaeologist
50 (1987) 118e126.
[61] A. Raban, in: G. Entjes-Nieborg, W. van Zeist (Eds.), Man’s role in the
shaping if the Eastern Mediterranean landscape, Balkema, Rotterdam,
1990.
[62] A. Raban, in: V. Karageorghis, D. Michaelides (Eds.), Proceedings of the
International Symposium ‘Cyprus and the Sea’, University of Cyprus,
Nicosia, 1995.
[63] E.G. Reinhardt, A. Raban, Destruction of Herod the Great’s harbor
Caesarea Maritima, Israel, geoarchaeological evidence, Geology 27
(1999) 811e814.
[64] E. Renan, La Mission de Phénicie, Imprimerie Nationale, Paris, 1864.
[65] K.W. Russell, The earthquake chronology of Palestine and Northwest
Arabia from the 2nd through the mid-8th century A.D, Bulletin of the
American School of Oriental Research 260 (1985) 37e59.
[66] G. Saadé, in: M. Yon M. Sznycer, P. Bordreuil, Le pays d’Ougarit autour
de 1200 av. J.-C. Ras Shamra-Ougarit, IX Actes du colloque internatio-
nalParis, 28 juine1er juillet 1993, Paris, 1995.
[67] R. Saidah, Fouilles de Sidon-Dakerman: l’agglomération chalcholithi-
que, Berytus 27 (1979) 29e55.
[68] M. Saghieh Beydoun, Urban planning in a seaport city: Beirut from Hel-
lenistic to Byzantine periods, Bulletin d’Archéologie et d’Architecture
Libanaises HSII (2005) 147e184.
[69] P. Sanlaville, Etude géomorphologique de la région littorale du Liban,
Publications de l’Université Libanaise, section des études géographiques,
Beirut, 1977.
[70] C.F.A. Schaeffer, Les fouilles de Ras Shamra-Ugarit, 6ème campagne
(printemps 1934), rapport sommaire, Syria 16 (1935) 141e176.
[71] S.L. Soloviev, O.N. Solovieva, C.N. Go, K.S. Kim, N.A. Shchetnikov,
Tsunamis in the Mediterranean Sea 2000 B.C.-2000 A.D., Kluwer
Academic Publishers, Dordrecht, 2000.
[72] D. Sourdel, J. Sourdel-Thomine, in: E. Poleggi (Ed.), Città portuali del
Mediterraneo: storia e archeologia, Sagep Editrice, Genova, 1989.
[73] J. Viret, Nouvelles données sur le port de Tripoli, Tempora, Annales
d’histoire et d’archéologie 10e11 (1999e2000) 117e138.
[74] J. Viret, in: F. Denise, L. Nordiguian (Eds.), Une aventure archéologique:
Antoine Poidebard, photographe et aviateur, Marseille, Arles, Beyrouth,
2004.
[75] J. Viret, Les ‘‘murs de mer’’ de la côte levantine, Méditerranée 104
(2005) 15e23.
[76] S. Wachsmann, Seagoing ships and seamanship in the Bronze Age
Levant, A&M University Press, Texas, 1998.