The Evolution of Akko Harbor and its Mediterranean ... · Bronze Age, Iron Age, and Persian Periods...

Journal of Island & Coastal Archaeology, 5:191–211, 2010Copyright © 2010 Taylor & Francis Group, LLCISSN: 1556-4894 print / 1556-1828 onlineDOI: 10.1080/15564891003664487

ARTICLE

The Evolution of Akko

Harbor and its

Mediterranean Maritime

Trade Links

Ehud Galili,1 Baruch Rosen,1 Dov Zviely,2 Na’ama Silberstein,1

and Gerald Finkielsztejn11Israel Antiquities Authority, Atlit, Israel2The Leon Recanati Institute for Maritime Studies, University of

Haifa, Haifa, Israel

ABSTRACT

Archaeological investigations carried out in Akko Harbor from 1992to 2004 are described and discussed, providing information on itslong history. During the Bronze Age, Iron Age, and Persian periods,maritime activity in Akko relied on a natural anchorage. Sedimentsand artifacts suggest that the harbor was first constructed during theHellenistic period and flourished since then. In the Byzantine period,the southern breakwater was in ruins and vessels anchored in the opensea. The exact location of the Early Islamic and Crusader harbors isunknown, but during the Crusades large vessels anchored in the opensea. The remnants of a fifteenth-century AD wooden pier indicate thatmaritime activity continued after the Crusader’s defeat. These changesare illustrated through the discussion of stratigraphy, the distribution ofarchaeological remains, and tectonic and sea-level considerations.

Keywords Israel, Hellenistic, Roman, crusade, Islamic

Received 15 Februbary 2009; accepted 3 August 2009.Address correspondence to Ehud Galili, Israel Antiquities Authority, P.O. Box 180, Atlit 30300, Israel.E-mail: [email protected]

191

Ehud Galili et al.

INTRODUCTION

The ancient cultural heritage of the Mediter-raneanencompassesnumerousreligions,cul-tures, and traditions, and reflects importantprocesses in the history of humanity. Duringthe middle Holocene the post-glacial sealevel rise reached its present level. At aboutthe same time the dawn of written historyoccurred, the large Near Eastern empiresbegan to consolidate, and permanent urbancenters appeared on the shores of the EastMediterranean. Intercultural trade intensifiedand included heavy and bulky items bettertransported by sea. The need arose for betterand bigger sea-going ships and these encour-aged the construction of effective shore facil-ities. As a result previously used natural shel-ters were being modified by ever intensifyinghuman actions. Additionally, sedimentationand erosion processes caused by natural andhuman activities modified the dynamics ofcoastal processes. Shore facilities stoppedfunctioningdue tosilting, as inKition,Cyprus(Mariner and Morhange 2007) and erosion asin Caesarea, Israel (Galili and Sharvit 1998),

and required extensive repairs and dredging.Oftentheshorelineshiftedseawardandmanyancient anchorages and harbors are nowinland. Understanding the history of shorefacility modifications is crucial to examiningthe development of marine civilizations.

Akko is a classical model of a fortifiedMediterranean coastal city and it has func-tioned for about 5,000 years as one of themain foci of maritime activity in the easternMediterranean. Underwater archaeologicalexcavations and surveys in Akko first aimedmainly to enable the dredging and construc-tion of a modern fishermen’s harbor in thewestern basin and later to protect the seafront. There are now plans for a new marinain the eastern basin (Figures 2 and 3) and thiscoastal development will significantly affectthe layout of the ancient harbor and impactarchaeological remains.

Excavations at Tel Akko, east of thepresent city, revealed imported artifacts andevidence for maritime trade from the MiddleBronzeAge(2200–1500BC)onward(Dothan1993). The finds indicate that the site musthave had a marine anchorage or harbor, but

Figure 1. Location map: a) the eastern Mediterranean Basin; b) insert of the northern Israeli Coast.

192 VOLUME 5 • ISSUE 2 • 2010

Evolution of Akko Harbor

Figure 2. Akko Port: insert a) seafront of the Pisan Quarter (detailed insert in Figures 14 and 15);b) Western Basin (detailed insert in Figures 3a and 4b) and possible locations of the EarlyIslamic chain blocking the harbor: 1) the western basin according to Linder and Raban(1965); 2) the eastern basin-according to Gertwagen (1996).

it is not known what the nature of thesefacilities was or where they were located. TelAkko was also occupied during the Iron Age(1200–587 BC) (Dotan 1993), but there is alack of archaeological evidence to determinethe level or location of maritime activity.

Until the present study it has not beenclear if the first harbor was constructed bythe Phoenicians during the Persian period(587–332 BC) or later during the Hellenisticperiod (333–37 BC). The importance of AkkoHarbor is documented by a long traditionof graphic depictions, starting with Romancoins; and from the fourteenth century bydrawings, charts, maps, and written docu-ments (Galili et al. 2004; Zviely et al. 2003).

However, the level of maritime activity dur-ing the Hellenistic, Roman (37 BC–AD 324),and Byzantine (AD 324–640) periods was notverified by underwater archaeological findsuntil recently.

The layout of the Early Islamic (AD740–1099) harbor and the locations of thebreakwater and the chain which blockedits entrance are not agreed upon. Someresearchers suggest that Akko Harbor wasbuilt as two basins similar to that of Tyre(Shur 1990). According to Linder and Raban(1965) and Jacoby (1979), the Early Islamicand Crusader harbors (AD 1099–1291) werein the western basin. It is unclear whetherthe chain blocked the western basin entrance

JOURNAL OF ISLAND & COASTAL ARCHAEOLOGY 193

Ehud Galili et al.

Figure 3. a) Location of ancient structures and shipwrecks in Akko Port: (1) late Roman/Byzantineshipwreck; (2) Byzantine shipwreck; (3) Crusader shipwreck; (4) hoard of Crusadersgold florins; (5) seventeenth- to eighteenth-century shipwreck hull; (6) eighteenth-centuryshipwreck hull with schist stones ballast; (7) hull section of an eighteenth-centuryshipwreck; (8 and 9) hulls of nineteenth-century shipwrecks; (10, 11, 12, 13, and 14)nineteenth- to early twentieth-century Ottoman shipwreck hulls; (15) wooden poles of athirteenth-century pier; (16) stone anchor; (17) mooring stone/anchor; (18) ancient dockmade of ashlars kurkar blocks; (19) a submerged foundation of the Tower of Flies built ofhuge headers; (22) Tower of Flies; 3b) the proposed marina layout.

(Flinderetal.1993; Jacoby1979;Raban1982)(Figure 2.1), or the eastern basin entrance(Gertwagen 1996) (Figure 2.2). The locationof an Early Islamic built harbor in the easternbasin is crucial for future planning of amodern marina in this place because state

planning policy in living historical cities isnot to exceed the layout of the ancientharbors. According to the designers of thenew marina, the ancient south breakwaterand the modern breakwater built on it do notprovide sufficient shelter for vessels in this

194 VOLUME 5 • ISSUE 2 • 2010


Figure 4. a) Distribution (%) of the Hellenistic, Roman, and Byzantine finds in the harbor; 4b)excavation areas.

basin (Raviv Rami, personal communication1998). To improve the protection offered bythe newly proposed marina in the easternbasin, there is a need to extend the southbreakwater to a distance of about 300 m tothe southeast (Figure 3b). This may suggestthat in the past the eastern basin could nothave provided sufficient shelter as expectedfrom a built harbor, but rather a temporaryanchorage. The existence of a separate har-bor of the Italian Pisa commune (so called‘Pisan Harbor’) during the Crusaders periodwas widely debated and the function of aCrusaders stone built tunnel crossing the cityfrom east to west is not yet understood.

Human activity on the shores and naturalcoastal changes due to sea level rise, seismicactivity, and tectonics are obliterating themaritime archaeological heritage. Massivecoastal developments are executed by pow-erful mechanical agents financed with im-mense capital. The predicted sea level rise inthe twenty-first century will also have crucialimpact on archaeological sites along theworld’s coastlines (Erlandson 2008). Urbancenters and living human societies can adapt,change, or move, but coastal archaeologicalsites cannot. To establish integrated and sus-tainable coastal zone management programs,there is a need to develop a proper archaeo-logical data base. There is a narrow window

to research, document, salvage, protect, andpreserve the coastal and marine culturalresources before further damage is done. Theancient city of Akko (Acre) is an ideal casestudy for the Mediterranean and may serve asa model for preservation and protection.

The present study summarizes and dis-cusses the results of archaeological investi-gations carried out by the Israel AntiquitiesAuthority (IAA) between 1990 and 2004.Detailed studies of numismatics, ceramics,fishing gear, and metal artifacts are beingpublished separately. Based on the resultsof these studies, we aim to clarify issuesconcerning the location and duration of theharbor and its history.

METHODS

In 1990–1991, archaeologists from theIAA conducted underwater archaeologicaland geophysical surveys in Akko Bay usingecho-sounder, sub-bottom profiler, andside-scan sonar to explore the sea floor of theharbor and nearby areas. In addition, waterjet drillings were carried out to evaluate thethickness and nature of the unconsolidatedsediments beneath the sea floor, to locateshipwrecks hidden in the sediments, and


Ehud Galili et al.

Figure 5. Bronze figurines (Persian period).

check anomalies and targets identified byremote sensing. In 1992–1993, the harborof Akko was dredged and archaeologicalmaterial was examined and studied by thestaff of the IAA. A crane lifted sedimentscontaining archaeological materials from theharbor bed and deposited them on a barge.The sediments were then taken to dumpingareas offshore, released, and examined bydivers. To map the source of the archaeo-logical artifacts, the harbor was divided intothree main areas (Figure 4b) using a grid of5 × 5 m squares. Although underwatersurveys were carried out in wide areas inand around the harbor, only the westernbasin and a small portion of the eastern basinwere dredged. Archaeological studies by

the IAA reported and discussed below werepublished in a series of preliminary reports(Galili et al. 1991; Galili and Rosen 2008a;Galili et al. 2002).

THE FINDS

Bronze Age, Iron Age, and Persian Periods

There are few finds assumed to be fromthe Bronze Age. These include two well-shaped stone anchors with one hole weigh-ing 39,280 kg (Figure 3a: no. 16, 17) and aloaf-shaped copper ingot weighing 4.5 kg.The dating of stone anchors is not alwaysaccurate, but scholars suggest that large and

196 VOLUME 5 • ISSUE 2 • 2010


well-shaped stone anchors with a single widehole (10–15 cm diameter) usually date to theLate Bronze Age (Frost 1963, 1970, 1973).An underwater survey in Neve-Yam, southof Atlit, ∼25 km south of Akko (Figure 1)(Galili and Sharvit 1999), exposed a LateBronze assemblage, including 83 loaf-shapedcopper ingots identical to the one fromAkko. This assemblage included hematiteweights in theshapeofwheatgrains, abronzeadze, a socketed spear head, bronze tongs,and several large (80–150 kg) stone anchorswith a single hole. The similarity of ingotsand anchors recovered from Neve Yam, theCape Gelidonia wreck (Bass 1961), and Akkosupport a Late Bronze Age dating of the latter.No artifacts from the Iron Age were foundat Akko and very few ceramic sherds andtwo bronze figurines from the Persian periodwere recovered (Figure 5).

Hellenistic Period

Most of the Hellenistic artifacts wererecovered from the western basin (Figures2b, 4b: areas D, E, F). Hellenistic artifactsinclude amphorae, amphorae handles withstamps, black ware vessels, black and redMegarian bowls, eastern Terra Sigilatta waretypes A, B, C, and D as well as western TerraSigilatta ware (Silberstein et al. in press).The large number of amphorae, domesticvessels (Figure 6), and metal objects, point toextensive trade relations with Mediterraneancountries and to large-scale imports from theAegean (Figure 7a).

Amphorae used in maritime trade forma significant portion of the ceramic assem-blage and originated in numerous sites inthe Mediterranean Basin. Of these identifiedas Hellenistic, about 80% originated in theAegean, 8% from the coasts of Syria andPalestine, 5% from Egypt and North Africa,and a few from Italy (Figure 7). This repre-sents clear evidence of extensive economicconnections with Greece during this time.About 50 stamped amphora handles wererecovered during the excavations in additionto numerous unstamped specimens (Finkiel-sztejn in press). As elsewhere in Israel, mostof the stamped handles originated in Rhodesand can be dated to the first half of the second

Figure 6. Hellenistic pottery vessel bearing in-scription “AKRATON” (young wine).

century BC; a few were from the second halfand the beginning of the first. More than halfdated to the years 188–166 BC (Finkielsztejnin press). The high number of handles foundmay be associated with military activity bythe Ptolemaic and the Seleucid armies, suchas the Battle of Banias (198 BC), where AkkoHarbor may have been used to supply thefighting force (Finkielsztejn in press).

Judging fromthewealthoffindsandtheirdiversity, maritime activities in the westernbasin of the harbor underwent a fundamentalchange during the Hellenistic period relativeto previous periods.

Roman Period

Most of the Roman material was recov-ered from the western basin (Figures 2b,4b: areas D, E, F). The finds include thewooden hull of one or more medium-sizedRoman/Byzantine ships (Figure 3a: no. 1).Timbers from the shipwreck were provision-ally 14C dated to the second–fifth centuriesAD (Kahanov Yaakov, personal communica-tion). Recoveries included a few coins, othermetal objects, and numerous whole and bro-ken ceramic vessels representing importedand local amphorae, cooking pots, oil lamps,bowls, plates, and Eastern Terra Sigillata


Ehud Galili et al.

Figure 7. Origin of the amphorae from Akko Port; a) percentage (%) of distribution and origin; b)location in the Mediterranean.

ware. Western Terra Sigillata ware, foundonly rarely on inland sites, were also found inthe harbor. Some of the pottery bears inscrip-tions and maritime themes (Figure 8). Halfof the identified Roman amphorae originatedfrom the Aegean and Black seas, 20% fromItaly and the western Mediterranean, 16%fromthecoastsofSyriaandPalestine,and10%from North Africa (Silberstein et al. in press)(Figure 7). The amphorae served as foodcontainers. The luxury pottery plates andwestern terra Sigillata vessels were imported

to provide for the needs of an emergingaffluent society.

Byzantine Period

A considerable amount of Byzantine pot-tery, a few pieces of wood, bronze nails,and a number of coins were found near theentrance of the harbor (Figure 4b: area C).Most of the amphorae belonged to threemain types and probably originated from

198 VOLUME 5 • ISSUE 2 • 2010


Figure 8. a) Terra sigillata plate with a depiction of a ship moored to a tree on the shore (a typicalnorth Mediterranean anchorage); b) Roman bowl decorated with iron anchors.

a single wreck (Figure 3a: no. 2). Nearlyhalf of the identified Byzantine amphoraeoriginated from the Aegean and Black seas,40% from the coasts of Syria and Palestine,3% from North Africa, and 6% from thewestern Mediterranean (Silberstein et al. inpress) (Figure 7). Eighth-century amphoraediscovered in the remains of the Byzantinewreck indicate that maritime trade by Byzan-tine sailors continued into the Early Islamicperiod. A few broken Byzantine iron anchorswere also recovered from the western basin(Figure 9). A concentration of 13 Byzantineiron anchors was observed on the sea bottom

some 2000 m southwest of the harbor,in 15 m of water (Kotzer Adam, personalcommunication 2004). They could representa cargo of anchors, which was on a ship thatwrecked on its way to the harbor.

Early Islamic Period

Very few finds from the Early Islamicperiod were found in the harbor—the onlydiagnostic pottery vessel recovered is onebowl base (Stern forthcoming). Howeverthe historian Al-Muqaddasi described the


Ehud Galili et al.

Figure 9. a) Broken Byzantine Iron anchor; b)a reconstruction of a Byzantine ironanchor.

construction of Akko Harbor by Ibn Tulunand its closure by a chain (Al-Muqaddasi1906:11, 162–163).

Crusaders Period

Sections of the wooden hull of a shipdated by 14C to AD 1062–1250 were discov-ered near the western basin entrance (Figure3a: no. 3). The remains include a keel anda few frames and planks partly covered with

materialusedasballast that resemblespouredcement; however, this requires further ex-amination. Crusader pottery vessels wererecovered mainly east of the entrance to thewestern basin (Figure 4b: areas A, B). Somesherds from this period were retrieved fromthe western basin, including fragments of 50bowls, 24 cooking pots, and 31 amphorae.Most of the bowls and the amphorae wereimported from the eastern Mediterranean.However, the cooking vessels demonstrate adifferent pattern, with most of them beinglocal and only 25% imported from Cyprus(Stern forthcoming). A hoard of 30 goldflorins minted in Florence in the secondhalf of the thirteenth century AD was recov-ered near the harbor entrance (Kool 2006)(Figures 10; 3a: no. 4). The coins show thatthe economic and political ascendancy ofFlorence had reached Akko by this time,presumably on board Pisan ships. Amongpossible explanations for the mishap is acritical moment in Akko‘s history. Historicaldocuments report that on May 18, 1291, aftera siege of several weeks, Frankish Akko fellto the advancing Mameluke army. Soldiersand citizens, desperate to escape the enemy,crowded into theharbor.Eye-witnesses, suchas the anonymous ‘Templar of Tyre” related

Figure 10. A hoard of Crusader gold florins recovered from area C.

200 VOLUME 5 • ISSUE 2 • 2010


Figure 11. Three Ottoman shipwrecks with wooden hulls, which will be covered by the new marinaproject (Figure 4: 11, 12, 13).

that a few noble ladies and merchants es-caped by bribing owners of small boats withjewelry and gold to be ferried offshore toships heading for Cyprus. Many, however,drowned with their precious possessions(Kool 2006).

Underwater surveys discovered twolarge (ca. 5 m long) iron grapnel anchors at adepth of 12 m, 1800 m south of the harbor.The anchors, about 5 m long laying about100 m apart with their tying rings pointingtowards the east and the northeast, probablybelonged to one or two large Crusader shipstrying to anchor in the open sea south of theharbor. Their position and the two missingbroken arms of the southern one couldindicate that during a storm they were lostby vessels which drifted to be wrecked onthe shore.

The Mameluke and Ottoman Periods

Several shipwrecks dating from the last300 years were discovered, including wholewooden hulls (Figures 3a: nos. 5, 6, 7, 8, 9,

10, 11, 12, 13, 14; 11). Numerous potteryitems were uncovered at the entrance tothe western basin and to the east (Figure4b: areas A, B). They include hundreds ofglazed ceramic bowls, tobacco pipes, do-mestic pottery vessels, iron cannonballs, andlead bullets. The ceramics point to traderelations with Turkey, Syria, Lebanon, andEgypt. Several ceramic vessels dated to thefourteenth–seventeenth centuries AD wereimports from Italy.

Near the entrance to the western basin,a group of 11 upright wooden piles withrectangular cross-sections was found (Figure12a, 12b). Additional piles belonging to thisconstruction were probably dislodged overthe years by fishermen’s nets, anchors, ordredging of the harbor entrance. Dating by14C placed them in the fifteenth century (cal.AD 1425–1460). The wood was identifiedas non-local European pine (Liphschitz Nili,personal communication 2002). This con-struction probably served as a base for aquay or a pier used for mooring large vessels(with a draught of 2 to 3 m) near the harbor


Ehud Galili et al.

Figure 12. a) Location of the thirteenth-century wooden pier; b) plan of the pier; c and d) possiblereconstructions of the wooden pier and its use.

entrance, but still in relatively deep water(Figure 12c, 12d). The structure may havebeen connected by a wooden bridge to theshallow rampart at the end of the southernbreakwater, or it may have been completelydetached from the shore.

DISCUSSION

Akko, a lively fishermen’s harbor today,exists in what is basically a medieval townwith an abundance of archaeological remainsthat should be protected and preserved.Our research here represents the long-termstudy of the harbor’s massive construction,earthworks, and development over the past20 years with the goal of recording the site’sunique historical context.

For the Bronze Age, Iron Age, and Persianperiods, Raban (1985) suggested that theNa’aman River served as an inland harborfor seagoing vessels. This idea has not yetbeen supported by archaeological or geo-morphological evidence. The Israeli river

mouths during these periods were shallowand, as today, blocked by sandbars andwere thus impassable for sea-going vessels(Galili 1986; Galili and Rosen 2008b; Poratet al. 2009; Zviely et al. 2006). Underwaterresearch in the last 40 years indicated thatBronze Age anchorages along the IsraeliCoast generally were located in areas partlyprotected by kurkar (calcareous sandstone)islands or sunken bars (Galili and Sharvit1994). Based on a pottery sherd bearing aPhoenician inscription (fifth to sixth cen-turies BC) found out of context to the westof the harbor, and the header structure ofthe breakwater, it was suggested that AkkoHarbor was built during the Persian period toaccommodate Cambyses‘s fleet on its way toEgypt (Linder and Raban 1965; Raban 1982,1983, 1993). Since no evidence for a harboror anchorage has ever been found in theNa’aman River channel, it is suggested thatthe anchorage during these periods was inthe lee of the promontory south of the citywhich was partly protected from the north

202 VOLUME 5 • ISSUE 2 • 2010


Figure 13. Schematic cross-section of the archaeological remains in Akko Port.

and north-west winds. The paucity of findsprior to the Hellenistic period suggests thatthe harbor had not yet been constructed(Figure 13). The absence of finds from theearly periods stands in contrast to the richfindings from archaeological excavations atTel Akko. This absence may be explained bythe intensive harbor building activities sincetheHellenisticperiodandmoderndeepeningand dredging operations. It is also possiblethat such remains are still buried under thicksediments on the harbor bottom. The latestdredging carried out during the current studydid not reach the kurkar bedrock as seen bythe dashed line marking the deepening area(Figure 13).

Two additional small Bronze Age anchor-ages which provided protection from thenorth winds were identified on the WesternGalilee Coast, north of Akko: one in Nahal(seasonal stream) Bet Ha-Emeq outlet (7 kmnorth of Akko) and the other south of Achziv(13.5 km north of Akko). In both anchorages,stone anchors with a single hole, generallyattributed to the Bronze Age, were found.It seems that in Akko and in the WesternGalilee there are no submerged kurkar reefs

close to the shore to provide shelter fromwestern and southwestern winds like thosefound in central and southern Israel. Thus,the anchorages on the Western Galilee Coast,including Akko, relied on east-west orientednatural raised rock features which pro-vided temporary shelter fromnorthernwindsonly.

The abundance of Hellenistic artifacts,embedded in fine silt sediment, attesting tolow energy waves and good mooring con-ditions in the western basin, indicates thatthe southern breakwater was built during theHellenistic period supporting the notion thatthe harbor was built at this time (Flinder et al.1993; Gertwagen 1996; Jacoby 1979). Themasonry header foundations and the earlystructures on the Tower of Flies were prob-ably also built during the Hellenistic period.The port of Tyre, which was a major East-ern Mediterranean harbor, was demolishedby Alexander in 332 BC, which may haveencouraged the proliferation of Akko harbor.

In the Roman period, Akko harbor wasthe main Roman sea gate to the land ofIsrael until the construction of Caesarea’sharbor by Herod. In 47/8 BC Julius Caesar


Ehud Galili et al.

landed in Akko on his way to Alexandria.The biography of Herod indicates that Akkoplayed an important role in his career. Ac-cording to Josephus, Herod landed therein February 39 BC after the Roman Senatedeclared him king and there he drafted anarmy for fighting Antigonus, his rival. Hemet Octavian (Augustus) in Akko during thesummer of 30 BC when the latter was on hisway to Egypt to fight Anthony and Cleopatra(Kashtan1988).Herodmusthavebeenawareof the critical role of Akko which was notincluded in his kingdom, as an access pointto Rome as well as the entire Mediterranean.This may have very well encouraged him tobuild Caesarea harbor. In AD 56, St. Paullanded in Akko on his way from Asia Minorand the port was a base for Vespasian duringthe Jewish wars (Kashtan 1988). Ceramicsfrom the harbor suggest that Akko enjoyedfar-flung trade relations with Mediterraneancountries in the first century AD. Luxuryimports reveal that the harbor of Akko serveda society that included wealthy consumers.Finds, including the hull of a medium-sizedship found inareaE, indicate thatmaritimeac-tivity was concentrated in the western basinwhich was protected by the southern break-water and provided shelter for vessels witha draught of up to 2.5–3 m. This agrees withnumismatic depictions of maritime symbolsand harbor constructions on Roman coinsminted at Akko, including an enclosed harborsurrounded by store houses or ship sheds andpossibly a lighthouse (Galili et al. 2004).

Archaeological materials dredged fromthe west basin bottom, at a depth of up toabout 4 m contained numerous Hellenisticand Roman artifacts, from the second centuryBC to the third century AD. It demonstratesthat at these periods, the western basin wasa flourishing port with a depth of at least 4m. The clay sediment in which the artifactswereembeddedtypifiesaclosedsedimentarybasin and indicates that it was protectedby an efficient south breakwater. The evi-dence shows wide commercial connectionsat this period with diverse Mediterraneanregions and port cities. The amphorae corpusindicates a significant rise in the amountof trade and demonstrates the unity of theMediterranean Basin under the rule of Rome.

For the Byzantine period, historicalsources testify that there was a shipyard inthe city, which was a regional center havingits own Bishop. The quantity of Byzantineartifacts from the western basin relative totheamountofRomanandHellenisticmaterial(Figure 2b), and the discovery of broken ironanchors indicate subsequent unfavorable an-choring conditions during the Byzantine pe-riod. Due to the lack of proper maintenanceafter the Roman period, the southern break-water probably ceased to protect seagoingcraft,compellingsailors toanchor intheopensea. The cargo of 13 iron anchors discoveredon the seabed southwest of the harbor mayindicate a demand for anchors needed duringthe Byzantine period for anchoring in theopen sea. The artifacts indicate that the maineconomic connections during the periodwere with the Byzantine Empire.

During the Early Islamic period:

Ibn Tulun saw the city of Tyre, itsfortifications and the wall surround-ing the port, and this is what hewanted to build in Acre. . . he hiredan old architect from Jerusalem,Abu Bakr. This architect orderedthe tying together of Sycamoretrunks, floated them on the wa-ter according to the land fortifi-cation dimensions, and tied themtogether, and built upon them withstones and mortar. Each five coursewere tied by long stones/pillars?After the courses sunk they wereleft on the sea bottom for a year inorder to stabilize. Then he returnedto build where he had stopped. (Al-Muqaddasi 1906:162–163)

At night, the opening of the harborwas closed by an iron chain (Al-Muqaddasi1906:162–163). To bridge so great a distance,the chain may have been suspended onfloated beams or wooden barges. A similartechnique was used several hundred yearslater by the Genoese who closed the harborin 1258 by a chain floating on wooden beams(Gertwagen 1996). It is uncertain whetherIbn Tulun harbor was in the eastern basinwhich offers limited protection from the

204 VOLUME 5 • ISSUE 2 • 2010


Figure 14. The seafront of the Pisan quarter, ruins of the Hellenistic-Roman breakwater, and asection of the seawall built on them (indicated by black arrow).

prevailing southwest storm winds. The na-ture of the eastern rampart was discussed byseveral archaeologists who have not yet suc-ceeded in locating the Early Islamic harborbuilt by Abu Bakr. Judging by Al-Muqaddasi’sdescription, the currently submergedeasternrampart, connecting the Tower of Flies tothe shore, was thought to have been built asa breakwater during this period (Gertwagen1996). However, underwater investigationsindicated that this structure is built fromloose rubble. No archaeological indicationsfor using the construction or the methodsdescribed by Al-Muqaddasi have yet beenfound in the eastern rampart. Raban (1982)suggested that this rampart was built as abarrier to block sediments from the Na’amanRiver and prevent the silting of the harbor.He suggested that the Early Islamic harborwas located in the western basin. Gertwagen(1996) suggested that the eastern rampartwas constructed as a breakwater to protectthe harbor against easterly winds. However,the proximity of the eastern shore of the bay

to the harbor prevents easterly winds fromraising high waves. Thus, the function of theeastern rampart as an effective breakwatermust be questioned. Given the archaeolog-ical evidence to date, and the location andlayout of the eastern rampart, it is proposedthat it was built as an approach road for thebuilding of structures on the Tower of Flies,and perhaps as a barrier preventing vesselsfrom approaching the harbor from the east.

During the Crusader period, Akko Har-bor was the country’s major port. The numer-ous military, economic, and civil activitiesin the harbor area are well documented.Thus, the scarcity of Crusader remains on theharbor bed was unexpected. New evidenceshows that during the Crusader period, thewestern basin was silted up due to neglectand poor maintenance, restricting most mar-itime activity in the bay to the partly exposedeastern basin. These conclusions supportJacoby‘s suggestion according to primarysource (Theuderic, ca. AD 1172) that duringthe Crusades the western basin was shallow


Ehud Galili et al.

Figure 15. a) Section of the sea wall built on the ruins of the ancient breakwater. The partlysubmerged arch (marked by white “a”), connecting two blocks (remains of theHellenistic/Roman breakwater and marked by white “b”). b) The sea wall after recentpreservation and protection works.

206 VOLUME 5 • ISSUE 2 • 2010


and was not used by deep-drafted vesselswhichanchoredintheopenseaortheeasternbasin (Jacoby 1979, 1985; Tobler 1865). Theeastern basin was not dredged during thepresent investigation, a fact that may explainthe paucity of artifacts from the Moslem andCrusader periods. Water jet drillings carriedout in the eastern basin, however, indicatethat the water depth in this basin couldhave reached up to 5 m. The delivery of alarge number of iron anchors from Venice toAkko, that were to be rented to mariners,was recorded in AD 1288 (Jacoby 1985).It is suggested that the demand for reliableanchors resulted from the inability to anchorin the protected area during moderatedsouthwestern storms. This forced ships toanchor in the open, unprotected, sea. Thetwo large iron grapnel anchors recoveredabout 1.8 km south of Akko may be aconfirmation of that. Large Crusader vesselshad to anchor in the open sea south of theharbor, where the rocky sea bottom ensuredproper anchor hold. An anchoring site inthis area is indicated on Akko Harbor plansfrom AD 1707 and on modern nautical charts(Galili et al. 2004; Zviely et al. 2003).

The existence of a separate Pisan Harborhad been assumed by several scholars. Ithas been proposed that part of the southseafront of the Pisan Quarter functioned asan external anchorage, and that there wasa gate to an inner Pisan Harbor, located inKhan Esh-Shuna (Raban 1982, 1993). Thepresent underwater surveys indicate that thesea wall was built on top of the ancientbreakwater remnants (Figures 14 and 15).These remnants and the later seawalls builton them blocked any potential entrance tothe Pisan quarter from the south. Thus therecould not have been an entrance to an innerharbor in this area (Kesten 1993).

Excavations in Akko in 1993–1994 ex-posed the arched ashlar tunnel leading fromKhan Al-Umdan, near the harbor, to thesupposed location of the Templar fortress onthe west shore of Akko (Figure 2, depictedin black lines). The tunnel is about 200 mlong and 3 m wide. About 100 m east of itswest end, it splits into two parallel tunnels,each 1.5 m wide (Avisar and Stern 1996).The floor of the tunnel is presently under

about 0.5 m of water. Bivalve shells attachedto the bottom of the channel walls indicatethat it was partly flooded by seawater orbrackish water, probably soon after it wasconstructed, and before it silted up. Severalvertical grooves on the walls probably servedto hold wooden partitions or cofferdams in-tended to cut off and seal portions of it. It wassuggested that the tunnel could have servedas a strategic subterranean passage (SternEliezer, personal communication). Normallyitmayhavebeenusedtoconveymaterials andpassengers from the harbor to the Templarfortress, as a sewer outlet, or both. Thelower part of the tunnel was excavated in thekurkar to a level ensuring partial flooding bygroundwater mixed with seawater on whichrafts or boats carrying goods may have beenfloated. The cofferdams enabled tunnelingand maintenance below sea level. Thus, theflooded tunnel floor indicates that duringthe Crusader period, sea level was similarto that at present. Underwater and coastalsurveys carried on the western seafront showthat the tunnel did not reach the sea at itswestern end (Sharvit and Galili 2002). Theuse of canals to convey goods and passengerswas already known in medieval Europe; theNaviglio Grande canal was built between AD1179 and 1209 to bring marble from LakeMaggiore to Milan Cathedral (Majo 1999).As such, the use of a hidden undergroundcanal supplying a medieval fortress througha hostile quarter is not surprising.

During the Mameluk and Ottoman peri-ods, iconographic evidence and archaeologi-cal finds indicate that large deep-draught ves-sels anchored in the partly protected easternbasin and merchandise was transported fromthe ship to shore by small boats. After AD1291, Akko was razed by the Mamelukes,and the remaining, already neglected, har-bor facilities were probably destroyed andblocked to prevent the return of Christians.Nevertheless, testimonies by travelers andpilgrims who visited the harbor of Akko indi-cate that despite the destruction, the harborcontinued to function during the fourteenthand fifteenth centuries (Shur 1990:161–163).The previously destroyed shrunken city con-tained warehouses of Venetian traders arriv-ing from Damascus who exported cotton


Ehud Galili et al.

to Europe through the harbor. The Post-Crusader Italian ceramics and ceramics tes-tifying to connections with Venice found inTel Yassaf, 4 km north of Akko (Stern 1999),may support the written sources and indicatethat cotton-trading activity from Akko harborcontinued after the Crusader period. Thewoodenpierdiscoveredclosetotheentranceto the western basin was probably builtin this period for the cotton trade and forloading and unloading local merchandise andpassengers. This pier probably replaced thefacilities destroyed by the Muslims and theneglected, non-functioning silted-up harbor.Similar installations, consisting of rows ofwoodenpillarsof similar sizeandshape,werefound in the northern bay of Atlit (Galili 1986;Rosen et al. 2004) and west of Tel Ashkelon.The Atlit structure was dated to the twelfthto thirteenth centuries AD and the one atAshkelon to the fifth–sixth centuries AD. Theconstruction of wooden marine installationsis an uncommon occurrence in the materialculture of the area, where large trees are rareand costly. In the early Islamic period, thelocal timberavailable formarineconstructionin Akko was sycamore which was useful asa temporary frame for stone constructionbecause it decayed fast under water, therebyallowing the stone-built superstructure tosettle into the seabed. For the very samereason it was useless for permanent woodenmarine structures. Evidence of importedtimber from Europe for building breakwaterswas uncovered in the Roman harbor atCaesarea and the Byzantine harbor at Tiberias(Brandon 1996). There the timber was usedin the construction of wooden forms, intowhich cement was poured. Evidence of aRomanperioddamconstructedof timberwasrecently uncovered in excavations at NahalTanninim north of Caesarea. Such buildingtechniques and construction materials wereapparently imported to the region fromEurope by the Romans and the Crusaders.But later, during the Islamic conquest, theuse of timber in marine construction de-creased and knowledge of the technologywas probably lost. Medieval constructionof wooden marine installations is knownfrom Alexandretta Bay, where a harbor calledPortusPalorum(PortoPali in Italian)wasbuilt

at ancient Mallus, near the trading city of Ayas(Italian Layazzo, now Yumurtalik)(Mutafian1993:119–126). Depictions of Akko harborin a drawing from AD 1686 and in maps fromlater periods indicate that at the end of theseventeenth century, large buildings, whichdid not survive long, were erected on thesouthern breakwater.

CONCLUSIONS

The archaeology and geology of Akko areaindicate that the region has been tectonicallystable for the last 2,200 years. No signifi-cant vertical tectonic or sea level changes(> ± 25 cm) have occurred in the area (Galiliand Sharvit 1998; Sivan and Galili 1999). Inthe southern seafront of the Pisan Quarter,an ashlar-built arch partially sunk in the sea isvisible in a section of the sea-wall that wasbuilt on the ancient breakwater remnants(Figure 15a). Previous researchers have usedthis arch as a sea level marker for Crusadertimes, suggesting that sea level was 1–2 mlower than today (Flemming et al. 1978;Gertwagen 1989, 1996; Neev et al. 1987).Recent underwater research indicates thatthis arch bridges a 2 m wide gap betweentwounderwaterblocks remaining fromanan-cient breakwater and that there is no ashlars-built opening under the arch (Figure 15b).Rather the arch supported the foundations ofthe sea-wall over the gap (Figure 14). Thusthis arch is not a good indicator for sea levelrise or tectonic subsidence.

During the Bronze Age, Iron Age, andPersian period, Akko Bay was probably usedas a natural anchorage. The harbor wasconstructed in the Hellenistic period. Thedistribution of finds reveals shifting of themarine activity from the western basin inthe earlier periods to the entrance of themodern harbor and outside it, to the par-tially protected eastern basin in the laterperiods. Most finds from the Hellenistic andRoman periods were recovered from thewestern basin. Numerous Byzantine artifactsrecovered at the entrance of this basin mayindicate intensive marine activity in this areaduring that period, or may come from a singleshipwreck. Crusader remains were retrieved

208 VOLUME 5 • ISSUE 2 • 2010


mainly near the entrance to the westernbasin. This distribution attests to fundamen-tal changes in the anchorage conditionsinside the harbor, with siltation preventingdeep-draught vessels from entering the west-ern basin. Archaeological finds indicate thatthe western basin reached the peak of itsprosperity during the Hellenistic and Romanperiods. The Byzantine period saw a wors-ening of the anchoring conditions, and thesouthern breakwater seems to have deterio-rated severely by that time. After the Byzan-tine period the western basin apparentlysilted up further so that only small vesselswere able to use it. The eastern rampartwas constructed from gravel and field stonesto allow passage to the Tower of Flies forconstruction purposes and possibly to hindervessels from approaching the harbor fromthe east. During the fifteenth century AD amooring facility using wooden pillars wasconstructed close to the entrance of thewestern basin. At the end of the seventeenthcentury CE, large buildings, which did notsurvive long, were erected on the southernbreakwater.

As noted in our introduction, Akko isa classic and historically significant fortifiedMediterranean coastal city. It has functionedfor roughly 5,000 years as one of the maincenters of maritime trade and other activitiesin the eastern Mediterranean. Archaeologicalstudies of the historical development of AkkoHarborprovideawealthof informationaboutthe history of the city and the maritime foun-dations of its economy. A planned coastaldevelopment in the port of Akko which willexceed the layout of the ancient harbor willinterfere with highly significant archaeolog-ical resources that should be preserved forfuture generations and further research.

ACKNOWLEDGEMENTS

We thank the Israel Antiquities Authoritywhich enabled this study; the archaeologistsJ. Sharvit, U. Dahari, E. J. Stern, R. Kool,Y. Kahanov, T. Shabi, E. Stern, and M.Aviam; Y. Arnon and the late A. Raban forhelping in the identification of the pottery;V. Fenwick, S. Arenson, J. Tresman, R.

Galili, and anonymous reviewers for theiruseful remarks; the editors S. M. Fitzpatrickand J. Erlandson for their comments andsuggestions; S. Ben-Yehuda and B. Galili forthe illustrations; T. Sagiv and J. Galili forphotography, and diver D. Moskovitch.

REFERENCES

Al-Muqaddasi, S. 1906. Ahsan at-Taqasum fiMa’rifat al-Aqalim. In Bibliotheca Geographo-rum Arabicorum (M. J. De Geoje, ed.):153.Leiden: E. J. Brill.

Avisar, M. and E. Stern. 1996. Akko, the OldCity. Hadashot Arkheologiyot 106:20–21 (inHebrew).

Bass, G. F. 1961. The Cape Gelidonya wreck:Preliminary report. American Journal of Ar-chaeology 65:267–276.

Brandon, C. 1996. Cements, concrete and settingbarges at Sebastos: Comparison with otherRoman harbor examples and the descriptionsof Vitruvius. In Caesarea Maritime: A Retro-spective after Two Millennia (A. Raban, K. G.Hulm, eds.):25–40. Leiden: Brill.

Dothan, M. 1993. Akko. In New Encyclopedia ofArchaeological Excavations in the Holy Land,Vol. 1. (E. Stern, ed.):17–24. Jerusalem: TheIsrael Exploration Society.

Erlandson, J. M. 2008. Racing a rising tide: Globalwarming, rising seas, and the erosion of humanhistory. Journal of Island and Coastal Archae-ology 3:167–169.

Finkielsztejn, J. in press. Amphora stamps fromAkko Harbour. IAA reports.

Flemming, N., A. Raban, and C. Goetschel. 1978.Tectonic and eustatic changes on the Mediter-ranean Coast of Israel in the last 9000 years. InProgress in Underwater Science, Vol. 3 (NewSeries), Proceedings of the 11th Symposiumof the Underwater Association (J. C. Gambleand R. A. Yorke, eds.):33–93. London: BritishMuseum of Natural History.

Flinder, A., E. Linder, and E. T. Hall. 1993. Surveyof the ancient harbour of Akko, 1964–1966. InStudies in Archaeology and History of AncientIsrael (M. Heltzer, A. Segal, and D. Kaufman,eds.):199–225. Haifa: Haifa University Press.

Frost, H. 1963. From rope to chain, on thedevelopment of anchors in the Mediterranean.Mariner’s Mirror 49:1–20.

Frost, H. 1970. Bronze-Age stone anchors fromthe eastern Mediterranean: Dating and identifi-cation. Mariner‘s Mirror 56:381–385.


Ehud Galili et al.

Frost, H. 1973. Anchors, the potsherds of marinearchaeology:Ontherecordingofpiercedstonesfrom the Mediterranean. In Marine Archae-ology (D. Blackman, ed.):397–410. London:Butterworths.

Galili, E. 1986. Historical sea level changes andseafaring along the southeastern MediterraneanCoast. Cities on the sea, past and present. InProceedings of the 1st International Sympo-sium on Harbours, City Ports and CoastalTopography (H. Kutiel, ed.):69–73. Haifa: Uni-versity of Haifa and the Technion-Israel Instituteof Technology.

Galili, E., U. Dahari, J. Sharvit, M. Aviam, and E.Stern. 1991. Marine Akko Survey, Final Report.Unpublished IAA Report, mimo. Jerusalem:IAA.

Galili, E. and B. Rosen. 2008a. Akko Harbour: Newfinds revealed while deepening the port. In NewEncyclopedia of Archaeological Excavationsin The Holy Land, Vol. 5. (E. Stern, H. Geva,and A. Paris, eds.):1558–1561. Jerusalem: IsraelExploration Society.

Galili, E. and B. Rosen. 2008b. Marine archaeologyin Israel—recent discoveries. In New Encyclo-pedia of Archaeological Excavations in TheHolyLand,Vol.5 (E. Stern,H.Geva, andA.Paris,eds.):1925–1934. Jerusalem: Israel ExplorationSociety.

Galili, E. and J. Sharvit. 1994. Classification of un-derwater archaeological sites along the Mediter-ranean Coast of Israel: Finds from underwaterand coastal archaeological research. In Actesdu Symposium International Thracia PonticaV: 1991 (C. Angelova, ed.):269–296. Sozopol,Bulgaria: Centre Darcheologie Subaqnatique.

Galili, E. and J. Sharvit. 1998. Ancient coastal instal-lations and the tectonic stability of the IsraeliCoast in historical times. In Coastal Tectonics(I. S. Stewart and C. Vita-Finzi, eds.):147–163.Special Publications 146. London: GeologicalSociety.

Galili, E. and J. Sharvit. 1999. Underwater survey intheMediterraneanSea1992–1996.Excavationsand Surveys in Israel 19:96–101.

Galili, E., J. Sharvit, U. Dahari, N. Bahat-Zilberstein,G. Finkielsztejn, E. J. Stern, R. Kool, Y. Kahanov,and B. Rosen. 2002. Akko Port, underwatersurveys. Hadashot Arkheologiyot 114:13–16and 12–15 (in Hebrew/English).

Galili, E., D. Zviely, and B. Rosen. 2004. Akko Portand its surroundings in coins and graphic depic-tions. Horizons in Geography 62:113–130 (inHebrew, English abstract).

Gertwagen, R. 1989. The southern shore of theOld City of Akko—‘The Pisan Harbour’. C.M.SNews 16:3–5. Haifa: University of Haifa.

Gertwagen, R. 1996. The Crusader port of Akko:Layoutandproblemsofmaintenance. InAutourde la Première Croisade, Actes du Colloquede la Society for the Study of the Crusadesand the Latin East, Clermont-Ferrand, 22–25juin 1995 (M. Balard, ed.):553–582. Paris:Publications de la Sorbonne.

Jacoby, D. 1979. Crusader Akko in the ThirteenthCentury: Urban Layout and Topography (Es-tratto dagli Studi Medievali, 3a Serie, XX, I).Spoleto, Italy: Centro Italiano Di Studi Sull’altomedioevo.

Jacoby, D. 1985. Venetian anchors for CrusaderAcre. The Mariners Mirror 71(1):5–12.

Kashtan, N. 1988. Akko- Ptolemais: A maritimemetropolis in the Hellenistic and Early Romantimes, 332 BCE–70 CE, as seen through theliterary sources. In Mediterranean Cities, His-toric Perspective (I. Malkin and L. Holfelder,eds.):37–53. Haifa: University of Haifa.

Kesten,A.1993.TheOldCityofAkko,Reconsider-ation,Criticisms.Akko, Israel:TheMunicipalityof Akko (in Hebrew).

Kool, R. 2006. A Thirteenth Century hoard of goldflorins from the Medieval harbour of Acre. TheNumismatic Chronicle 166:301–320, plates57–61.

Linder, E. and A. Raban. 1965. Underwater surveyof Akko Harbour. In The Western Galileeand the Galilee Coast. The 19th Conven-tion for Yediat Haaretz. Jerusalem:180–193 (inHebrew).

Majo, A. 1999. II Duomo di Milano: Storia, Arte,Vita Religiosa. Milano: San Paolo Edizioni.

Marriner N. and Morhange C. 2007. Geoscience ofancient Mediterranean harbours. Earth-ScienceReviews 80:137–194.

Mutafian, C. 1993. Le Royaume Armenien deCilicie, XIIe-XIVe Siècle. Paris: CNRS.

Neev, D., N. Bakler, and K. O. Emery. 1987.MediterraneanCoastsof IsraelandSinai.NewYork: Taylor and Francis.

Porat, N., D. Sivan, and D. Zviely. 2008. LateHolocene embayment and sedimentologicalinfill processes in Haifa Bay, SE Mediterranean.Israel Journal of Earth Sciences 57:21–31.

Raban, A. 1983. Akko Harbour. Kardom24–25:141–146 (in Hebrew).

Raban, A. 1982. The harbours of Akko. In TheAntiquities of the Western Galilee (M. Yeda’aia,ed.):180–207. Tel-Aviv: Hakibutz Hamehukad(in Hebrew).

Raban, A. 1985. The ancient harbours of Israelin biblical times. In Harbour Archaeology,Proceedings of the First International Work-shop on Ancient Mediterranean Harbours,Caesarea Maritima (A. Raban, ed.):11–44.

210 VOLUME 5 • ISSUE 2 • 2010


Publication No. 1. Haifa: Haifa University; Ox-ford: BAR International Series 257.

Raban, A. 1993. Maritime Akko. In New Encyclo-pedia of Archaeological Excavations in theHoly Land, Vol. 1 (E. Stern, A. Lewinson-Gilboa,and J. Aviram, eds.):29–31. Jerusalem: IsraelExploration Society.

Rosen, B., R. Zvitov, E. Galili, and A. Nussinovitch.2004. Copper preservation of pinewood sub-merged for 550 years. Michmanim 18:9–18,79–83. Haifa: Haifa University.

Sharvit, J. and E. Galili. 2002. Akko west:Coastal and underwater surveys. HadashotArkheologiyot 114:11–13 and 10–12 (inHebrew/English).

Shur, N. 1990. The History of Akko. Tel-Aviv: Dvir(in Hebrew).

Silberstein, N., E. Galili, and J. Sharvit. in press. TheHellenistic, Roman and Byzantine pottery fromAkko Harbour. IAA reports, Jerusalem.

Sivan, D. and E. Galili. 1999. Holocene tectonicactivity in the Galilee Coast and shallow shelf,

Israel: A geological and archaeological study.Israel Journal of Earth Sciences 48:47–61.

Stern, E. J. 1999. The pottery of thethirteenth–fifteenth centuries from Giv’at Yasaf(Tel Er-Ras). Atiqot 37:174.

Stern, E. J. in press. The pottery of the Fatimid,Crusader, Mamluk and Early Ottoman Periodsfrom the deepening of the ‘Akko Marina’.Atiqot, Jerusalem.

Tobler, T. (ed.) 1865. Theuderic Libellus de LocisSanctis Editus Circa A.D. 1172. Paris: St. Gallen-Paris.

Zviely, D., E. Galili, and B. Rosen. 2003. AkkoPort in modern nautical maps. Horizons inGeography 56:62–77 (in Hebrew, Englishabstract).

Zviely, D., D. Sivan, A. Ecker, N. Bakler, V.Rohrlich, E. Galili E. Boaretto, M. Klein, andE. Kit. 2006. The Holocene evolution of HaifaBay area, Israel, and its influence on the ancienthuman settlements. The Holocene 16(6):849–861.


The Evolution of Akko Harbor and its Mediterranean ... · Bronze Age, Iron Age, and Persian Periods...

Documents

Transcript of The Evolution of Akko Harbor and its Mediterranean ... · Bronze Age, Iron Age, and Persian Periods...