Quaternary International 602 (2021) 66–81 Contents lists available at ScienceDirect Quaternary International journal homepage: www.elsevier.com/locate/quaint Using a multi-proxy approach to locate the elusive Phoenician/Persian anchorage of Tel Akko (Israel) Matthieu Giaime a, b, c, *, Harry M. Jol d, Yossi Salmon b, c, Gloria I. López b, c, Amani abu Hamid b, c, e, Logan Bergevin d, Paul Bauman f, Alastair McClymont f, Ethan Sailer-Haugland d, Michal Artzy b, c a Institut de Ciència I Tecnologia Ambientals (ICTA-UAB), Universitat Autònoma de Barcelona, 08193 Cerdanyola Del Vallès, Barcelona, Spain b Hatter Laboratory, Leon Recanati Institute for Maritime Studies, University of Haifa, Abba Khoushy Ave 199, Haifa, 3498838, Israel c Department of Maritime Civilizations, The Leon. H. Charney School of Marine Sciences, University of Haifa, Abba Khoushy Ave 199, Haifa, 3498838, Israel d Department of Geography & Anthropology, University of Wisconsin-Eau Claire, 105 Garfield Avenue, Eau Claire, WI, 54701, USA e Israel Antiquities Authority, Rockefeller Archeological Museum, P.O.B 586, Jerusalem, 9100402, Israel f BGC Engineering INC. Suite 500 - 1000 Centre St. NE, Calgary, AB, T2E 7W6, Canada A R T I C L E I N F O A B S T R A C T Keywords: Previous geoarchaeological research on the Akko coastal plain have contributed to the understanding of the Anchorages/harbors ancient coastal interface and added evidence as to the location/shift of the ancient anchorages dating from the Geoarchaeology Middle Bronze Age (beginning of the 2nd Millennium BC) to the Early Hellenistic period (mid of the 2nd century Geomorphology BC) of the ancient site of Tel Akko. The present research provides new insights into the environmental changes Ostracods and likely anchorage sites along the western edge of Tel Akko in the 1st Millennium BC (Iron Age II and III, Sedimentology Electrical resistivity tomography (ERT) periods associated with the Phoenician mariners and Persian army incursion). Our approach for locating the Ground penetrating radar (GPR) anchorage is based on a detailed investigation of subsurface sediments combining sedimentological and faunal analysis and radiocarbon dating of cores as well as identification of ceramic sherds found in the cores, and ground penetrating radar (GPR) and electrical resistivity tomography (ERT) surveys. Paleoenvironmental changes are compared and contrasted with the results of the archaeological investigations on the tell and in its vicinity. Our new data demonstrates that the Phoenician/Persian maritime interface of Tel Akko was mainly oriented toward the southwestern area of the tell where a natural anchorage was likely to have been located. At that time, the water depth in this area was ca. 2m, allowing for the anchorage of seagoing vessels. Increasing sediment deposition lead to the deterioration of direct, and eventual loss of access, to the sea. These conditions initiated the abandonment of the tell in the Early Hellenistic period as well as the westward shift to habitation on the peninsula, now the ‘Old city of Akko’, the Crusaders’ Saint Jean d’Acre. 1. Introduction (Giaime et al., 2019). In the 3rd and 2nd Millennia BC, the eastern Mediterranean populations favoured natural coastal features as The layout of ancient coastal cities is dictated by the shoreline po­ anchorage areas whereas artificial harbour basins and related port fa­ sition as well as any neighbouring river. In dynamic geomorphic set­ cilities started to be developed mainly during the Iron Age (Marriner tings, palaeoenvironmental reconstruction of ancient coastal et al., 2014, 2017). Along the coast of the southern Levant, the lack of settlements are particularly challenging unless a multi-disciplinary bays on the coast, encouraged the utilizations of river-mouths for approach is taken. The natural anchorages or artificial harbours – life­ maritime activities during the Bronze Age (Raban, 1985, Fig. 1A). lines of ancient coastal sites – may evolve or be relocated over time. This River-mouths have generally been used as safe havens for relatively depends on the change in orientation of the coastline due to sea-level short time periods. Over the long term they were vulnerable to a changes, sedimentary inputs, coastal erosion and associated hazards multitude of natural hazards, which include infilling by both fluvial and * Corresponding author. Institut de Ciència I Tecnologia Ambientals (ICTA-UAB), Universitat Autònoma de Barcelona, 08193 Cerdanyola Del Vallès, Barcelona, Spain. E-mail address: matthieu.giaime@gmail.com (M. Giaime). https://doi.org/10.1016/j.quaint.2021.06.008 Received 31 August 2020; Received in revised form 31 May 2021; Accepted 6 June 2021 Available online 23 June 2021 1040-6182/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). M. Giaime et al. Quaternary International 602 (2021) 66–81 Table 1 Main historical periods of Tel Akko from the Bronze Age to the Hellenistic period. Period Years BC Early Bronze 3500–2200 Middle Bronze 2200–1550 Late Bronze 1550–1200 Iron Age I 1200–1000 Iron Age II (Phoenician) 1000–586 Iron Age III (Persian) 586–332 Hellenistic 332–37 Beeri, 2008, Fig. 3). The development of this urban centre deeply transformed the pre-existing forest ecosystems toward an urban envi­ ronment with a vegetation biome dominated by cultivated species and urban weeds (Kaniewski et al., 2013). Written sources, including letters from the rulers of Akko, found mainly in Egypt, identifies the city as an important harbour town serving the Egyptian authorities in the mid-second Millennium BC (Artzy, 2012), although excavations have only located the latest part of the Late Bronze (late second Millennium BC) until now. The second period of economic growth at Tel Akko is centred on the second part of the Iron Age including the Persian Period at which time, signs of iron works were noted. The habitation pattern and the popu­ lation density increased after ca. 700-600 years BC (Artzy and Be’eri, 2010). The importance of Tel Akko in the trading network of the Levantine coast during the Iron Age is attested by the presence of Phoenicians, Persians and Greeks merchants (Gambash, 2014). It wasn’t until the Hellenistic period (3rd and early 2nd century BC) that the centre of the town moved to the promontory, where the ‘Old City of Akko/Acre is located (Artzy, 2015, Fig. 2). By using a multidisciplinary geo-environmental approach, we aim to present the geomorphological setting of the western area of Tel Akko in the Iron Age where a marine environment has been evidenced in our previous research (Morhange et al., 2016; Giaime et al., 2018). We seek to determine the possibilities of an anchorage and associated environ­ mental limitations for that area. The reconstruction of environmental changes affecting the western area of Tel Akko is fundamental for the understanding of the development of the tell as well as the evolution of the habitation patterns during the Phoenician/Persian period. 2. In search of Tel Akko’s Phoenician/Persian anchorage: (geo) archaeological context The coastal site of Tel Akko was founded on the northern limit of the Haifa/Akko Bay atop a naturally-occurring sand and kurkar feature. At Fig. 1. A) General location of Haifa Bay in the southern Levant (after Inman, Tel Akko, the kurkar, a local eolianite, is part of the Kurdane formation 2003; image ©CNES). (Kafri and Ecker, 1964) which is pre-Pleistocene of age (Sivan et al., Dominant wind and wave patterns at Alexandria are shown in the rose dia­ 1999). The site is located in a dynamic geomorphological environment grams (Stanley and Bernhardt, 2010). B) Geomorphological reconstruction of Haifa Bay 4000 years BP (after Elyashiv et al., 2016). at the end of the Nile’s littoral cell and in the proximity of the Na’aman River (Belos) and its mouth (Morhange et al., 2016; Giaime et al., 2018, Fig. 1B). Haifa/Akko bay was formed by the subsidence of a geological marine sediments due to sea level stabilisation (Giaime et al., 2019). graben delimited by two normal faults (Fig. 1B). It has undergone sig­ The UNESCO World Heritage site of Akko/Acre is one of the oldest nificant landscape changes during the Holocene (Zviely et al., 2006; continuously inhabited sites along the Levantine coast although its Elyashiv et al., 2016, Fig. 1B). Following postglacial sea-level rise and habitation pattern has evolved through time (general chronology given subsequent flooding of the graben, the shoreline transgressed several in Table 1). The main change occurred at the beginning of the Hellenistic kilometres to the east, reaching its maximum landward position ca. 4000 period when the tell was deserted in favour of the Akko promontory years ago (Zviely et al., 2006). During the past 4000 years, the coastline where the city is presently situated (Artzy, 2015, Fig. 2). rapidly prograded to the west. The main driver of this progradation was The first archaeological remains at Tel Akko are dated to the the deposition of 700 million m3 of Nile-derived quartz-rich sands beginning of the Early Bronze Age, ca. 5000 years ago. The first phase of transported northerly by longshore currents (Zviely et al., 2007) while urbanisation took place on the tell at the beginning of the Middle fluvial sediment inputs from the local rivers appears as a secondary Bronze. An impressive rampart reaching a height of 22 m and a width of factor with exceptions for the northern and southern extremities of the 60 m dated to the Middle Bronze IIa period was erected on the northern bay where the two main coastal streams flow into the sea (Vachtman side of the tell. The highest part of it was located in Area AB, where a et al., 2013; Zviely et al., 2006). large MBII building was constructed (Dothan, 1976; Raban, 1983, 1991; An artificial ‘Phoenician-style’ harbour was identified following 67 M. Giaime et al. Quaternary International 602 (2021) 66–81 Fig. 2. Location of the main archaeological areas and main habitation centres cited in the text. Background image © Zoom Earth, Microsoft. Fig. 3. A) Location map of ERT lines collected on Tel Akko southern plain (background image taken by drone © P. Bauman, 2018). B) Location map of the sedi­ mentary cores, GPR grids (to add), ERT profiles and excavated areas at Tel Akko. Start of the ERT line (0m) is shown by the yellow dot (background image taken by drone © J. Quatermaine, 2019). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) submarine archaeological prospections around the ‘Tower of flies’ re-study of the ceramics from Raban’s excavations, have not shown any which is still a landmark feature in the Bay (Linder and Raban, 1965; pre-Hellenistic ceramic evidence. In addition, archaeological excava­ Raban and Stieglitz, 1993, Fig. 2). It has been dated to the 6th century tions in the Akko promontory where the ‘Old City’ is located have not BC, the time of the of the Persian king, Cambysis’ incursion during his shed light on pre-Hellenistic occupation. Moreover, dredging of Akko’s conquest of Egypt. However, more recent studies have qualified this fishermen’s harbour has not provided evidence for the presence of dating. Artzy (2012) proposes that ‘Phoenician-type’ harbours pre-Hellenistic material in the harbour basin (Galili et al., 2010; Galili continued to be constructed in the south-eastern Mediterranean well and Rosen, 2017). Gambash (2014) states that during the mid 6th cen­ after the Phoenician Period. One example is the harbour at Amathus, tury BC of the Persian period, Akko did not serve as a logistical centre Cyprus (Empereur and Verlinder, 1987; Empereur, 1995; Empereur and regular point d’appui for Persian armies and fleets. It was only in the et al., 2018) which, despite depicting some Hellenic elements, was still late 5th and 4th centuries BC that the Persian gathered their army of constructed mainly in the Phoenician style around 300 BC. In addition, a mercenaries at Tel Akko under Artaxerses II and III (Gambash, 2014; 68 M. Giaime et al. Quaternary International 602 (2021) 66–81 Diodorus Siculus XV.41). resolution of the data (smaller spacing results in higher horizontal and On Tel Akko, archaeological finds support the fact that the tell was vertical resolution). Similarly, the length of the array controls the depth the main centre of the area during the Persian period. On the summit of of investigation (longer arrays yield greater investigation depths). Data the tell (Area A, Fig. 3), a Phoenician ostracon (a potsherd used as a collection was carried out in a sequential and automated fashion that writing surface), the longest ever found in Phoenicia, was unearthed takes advantage of all possible combinations of current injection and (Dothan, 1985a). More Phoenician ostraca, not yet published, were potential measurement electrodes. The data are downloaded to a com­ found in Area G, on the western side of the tell (Fig. 3), where a massive puter for processing and analysis. The data were then inverted using a building was noted as well (Artzy and Quatermaine, 2014). The latest two-dimensional (2-D) finite difference or finite element inversion coins found in Area G are dated to the first half of the second century BC routine using the RES2DINV software. To enhance layer boundaries at (Artzy, 2012). Rich remains from the Late Persian Period (late 5th-4th shallow depths, the 2011 ERT profiles were inverted using a robust (L1- century BC) as well as numerous imports from the Aegean world were norm) constraint on the model. Conversely, due to lower resistivity found during excavations (Artzy and Be’eri, 2010; Dothan, 1976, contrasts between subsurface layers surrounding Tel Akko, a smooth 1985b), especially in Area F in the north-western side of the tell (Raban, model (L2-norm) constraint was used for the 2018 ERT profiles. The 1993, Fig. 3). Area T, also on the western side of the tell, was studied model output from each ERT line is a 2-D cross-section plotting re­ during a salvage excavation in 2010 directed by Abu Hamid of the Israel sistivity (in ohm-m), versus depth. All ERT sections are presented as Antiquities Authority (IAA). Results showed the presence of a small in­ colour grids with warm colours (reds and pinks) indicating relatively dustrial area dated to the 4th century BC, where Late Persian to Early high resistivities and cool colours (light and dark blues) showing regions Hellenistic ceramics were found (Artzy, 2012, Fig. 3), as well as a of low resistivities. gravelly-sand facies bearing shells and ceramics interpreted as a high-energy coastline (Artzy, 2012). Pit surveys carried out, as part of 3.2. Ground penetrating radar (GPR) the ‘Total Archaeology’ Project, directed by Killebrew and Artzy since 2010, accentuate the extent of 5th-4th century BC finds including an GPR can be used to investigate and identify human-made and/or unusual large numbers of sherds, especially of Maritime Transport geomorphic features based on the sediments’ dielectric permittivity and Containers, both local and imported. electrical conductivity (Conyers and Goodman, 1997). GPR is a Based on the above evidence, Tel Akko was a major commercial hub non-invasive near-surface geophysical surveying technique that uses of the Mediterranean trade network in the Persian period, so it is hard to electromagnetic waves to image the subsurface (Jol and Bristow, 2003). assume that its harbour was situated in the bay, 1.5 km from the main A GPR survey was collected using a Sensors & Software pulseEKKO habitation and storage area. 1000 GPR system with an antennae frequency of 225 MHz (0.1 m step size and 0.5 m antenna separation). The grid dimensions were 24 m × 3. Material and methods 10.75 m (individual lines are 0.25 m apart). Site determination was based on archaeological excavations (Td) on the tell. Over the past ten years, geoarchaeological studies have shed light on After grid collection occurred, GPR lines were imported into the importance of fluvio-coastal changes and possible location of Tel GFP_Edit 4 (GPR Files and Parameters) and an axis-based grid was Akko’s anchorage(s) as early as the Middle Bronze Age (Morhange et al., created. The created GFP file is then exported to a Sensors and Software 2016; Giaime et al., 2018; Artzy et al., 2021). The focus of this paper lays GPR processing application - EKKO_Project 5. GPR profiles were pro­ on the geomorphological evolution and the possibilities of anchorage on cessed in a software extension called SliceView (gain adjusted between 7 the western edge of the tell (Figs. 2 and 3). We carried out a detailed and 8). Conductive sediments affected the average velocity for the grid sub-surface multi-disciplinary investigation. Ground penetrating radar and was calculated to be between 0.07 and 0.08 m/ns based on hyper­ (GPR) and electrical resistivity tomography (ERT) surveys combined bolic measurements. with sedimentological and faunal analysis of sedimentary cores. Cores were radiocarbon dated and ceramic sherds found in the cores were 3.3. Collection and analysis of the cores analysed to provide high-precision relative chronology. Sediment cores, GPR grids and ERT lines have been geographically located with a Dif­ Ten sediment cores were collected in plastic liners using a hydraulic ferential Global Positioning System (DGPS) and altitude benchmarked percussion driller (Geoprobe DT22). Prior to split-opening and logging relative to present mean sea level (msl) with a Total Station. the sediment cores, every core section was initially run through a Multi- Sensor Core Logger (MSCL) (PetroLab facility, University of Haifa). 3.1. Electrical resistivity tomography (ERT) Magnetic susceptibility (MS) measurements were key in deciphering between “natural” sedimentation (i.e., weak or negative MS values The ERT profiles were collected in two phases: ERT profiles Akko-1, produced by diamagnetic properties of well sorted quartz sand layers; Akko-2, and Akko-3 were acquired in July 2011, and ERT profiles Akko- Thompson and Oldfield, 1986) and more anthropogenic-like signatures 14, Akko-16, and Akko-17 were acquired in February 2018. The ERT (characterised by fluctuating MS values mainly due to the presence of profiles collected in 2011 were designed to investigate targets on the ceramics or changes in sediment texture). Core sub-samples from every flanks of Tel Akko to depths of up to 20 m below ground surface (mbgs), stratigraphic unit were taken for micropaleontological and sedimento­ and were acquired using an ABEM SAS 3000 resistivity system, a min­ logical analyses. General sediment texture, i.e., gravel (>2 mm), sand imum electrode spacing of 1.5 m, and with standard Wenner array (63 μm–2 mm) and silty-clay fractions (<63 μm), was determined by acquisition sequences. The 2018 profiles were designed to investigate wet-sieving. Ostracods were picked from the sand fraction >125 μm and deeper geological layers surrounding Tel Akko and were acquired with a identified using atlases and specific literature (Athersuch et al., 1989; multichannel ABEM Terrameter LS unit, a minimum electrode spacing of Lachenal, 1989; Frenzel and Boomer, 2005; Avnaim-Katav et al., 2016). 5.0 m, and expanded gradient array acquisition sequences. Positions and Molluscs were identified only from the gravel fraction (D’Angelo and topographic profiles for each ERT line were surveyed using handheld Garguillo, 1978; Avnaim-Katav et al., 2016). Trimble Geo XH GPS units, with a nominal accuracy in vertical and horizontal directions of less than 1 m. 3.4. Chronology For the 2011 and 2018 ERT surveys, resistivity data were collected through a linear array of electrodes coupled to a direct current (DC) Four samples (grape seeds and charcoals; Table 2) were radiocarbon resistivity transmitter and receiver, and an electronic switching box. The dated by Accelerator Mass Spectrometry (AMS) at the Radiochronology spacing between electrodes largely controls the horizontal and vertical Laboratory of the Centre d’études nordiques (University of Laval, Canada) 69 M. Giaime et al. Quaternary International 602 (2021) 66–81 Table 2 AMS-14C data expressed in calibrated years BP and BC at the 95% confidence level (2σ). using Calib 8.2 (Stuiver and Reimer, 1993; Stuiver et al., 2020) and the IntCal20 curve (Reimer et al., 2020). amsl: above mean sea level. Sample Lab Material Depth (cm below Elevation (cm Age 14C 2 sigma cal. BP min; 2 sigma cal. BC/AD min; Historical number surface) amsl) max max Period AK-XIX- ULA-8700 Charcoal 578 45 amsl 2800 ± 2850; 2960 1010; 900 BC Early Iron Age 3578 20 AK-XIX- ULA-8695 Charcoal 565 106 amsl 2455 ± 2370; 2700 750; 420 BC Late IA - 4565 15 Persian AK-XIX- ULA-8698 Charcoal 678 141 amsl 2455 ± 2370; 2700 750; 420 BC Late IA - 5678 20 Persian AK-XIX- ULA-8697 Seed 710 106 amsl 2485 ± 2470; 2720 770; 520 BC Late IA - 5710 (grape) 20 Persian and the Keck Carbon Cycle AMS Facility (University of Cal­ 4. Results ifornia–Irvine, USA). Identification of several ceramic fragments recovered from the cores provides a high-precision relative chronology 4.1. ERT measurements for specific stratigraphic units (Table 3). ERT lines Akko-1, Akko-2 and Akko-3 were surveyed on the western slope of the tell (Figs. 3 and 4). They demonstrate the presence of a Fig. 4. ERT lines Akko-1, Akko-2 and Akko-3. The profiles were collected on the western edge of Tel Akko (see Fig. 3). Horizontal scale is 9.42 pixels per unit spacing. Vertical exaggeration = 1.23. High resistivity features attest of the presence of sands/kurkar sandstone. 70 M. Giaime et al. Quaternary International 602 (2021) 66–81 relatively low to moderate resistive material (blue to green colours in Discrete, moderately resistive features are present at 10–20m depth. the image) in the first 4–6 m below surface and interpreted as fine- Only the northern part of ERT lines Akko-14 and Akko-16 shows high grained sediments such as an anthropogenic/agricultural soil. Highly resistive material (pink colour) interpreted to be dry soil and colluvium resistive features (>200 Ω-m; red to pink colours), interpreted as a from the tell, overlying kurkar sandstone (Fig. 5) although it is chal­ kurkar outcrop or sandy unit (both exhibit the same high resistivity lenging to differentiate the nature of the subsurface sediments in this values in the ERT profiles) are shown deeper in the subsurface. area as kurkar, dry sands and colluvium will exhibit the same high re­ Three 450-500 m-long ERT lines were surveyed in order to map the sistivity values (see Fig. 6). subsurface sedimentary sequence of the coastal plain. Lines Akko-14 and Akko-16 have a mostly north/south orientation, whereas Akko-17 trends approximately west/east (Figs. 3 and 5). Most of the sections 4.2. GPR are dominated by a low resistivity (blue colour) material interpreted as fine-grained (fine sands and silty-clays) over-saturated sediments. Ground penetrating radar data was collected on the western portion of the tell (near Td excavation) where a coastal environment was Fig. 5. ERT lines Akko-14, Akko-16 and Akko-17 collected on the Tel Akko coastal plain (see Fig. 3). Most of the sections are dominated by low resistivity (blue) material which are interpreted as fine over saturated sediments (silty-clay and fine sands). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) 71 M. Giaime et al. Quaternary International 602 (2021) 66–81 Fig. 6. Ground penetrating radar data was collected on the western portion of the tell (near Td excavation). The upper plan view image is a depth slice from 0.44 m to 0.55 m depth. A curvilinear feature is highlighted in this depth horizon. The lower image is Line 9 of the data collected. From 0 m to 12 m, multiple subhorizontal to horizontal contin­ uous reflections become dipping reflections at 12 m–20 m. The reflection pattens from the GPR data collected are interpreted as an aggradational and progradational geomorphic environ­ ment such as a coastline. interpreted based on archaeological finds. The upper planview image is 4.3. Core analyses a depth slice from between 0.44 m and 0.55 m depth. A curvilinear feature is highlighted in this depth horizon. The lower image is indi­ Stratigraphic units were described for each of the 10 cores analysed vidual Line 9 of the grid data. From 0 m to 12 m, multiple subhorizontal for this study according to sedimentary texture and colour, MS proper­ to horizontal continuous reflections become dipping reflections from 12 ties and eventually faunal content. The environment identified for each m to 20 m. The reflection pattens discovered within the GPR data stratigraphic unit is interpreted according to the Mediterranean benthic collected are interpreted as an aggradational and progradational classification system (Pérès and Picard, 1964; Pérès, 1982; Figs.1 Sup­ geomorphic environment such as a coastline. plementary Material) Analysis of the ceramic sherds found in the cores confirms the robustness of the radiocarbon chronology measurements, however, radiocarbon dating of one piece of charcoal shows a slight discrepancy that could be attributed to the old wood effect or the 72 M. Giaime et al. Quaternary International 602 (2021) 66–81 reworking of older material (Sample ULA-8700; Table 2). 4.3.2. Cores AK-XIX-6 and AK-XIX-7 AK-XIX-6, drilled at an elevation of 6.80 m amsl and AK-XIX-7 (7.33 4.3.1. Core AK-XIX-8 m amsl) are the northernmost cores of this study (Figs. 3 and 8). The core Core AK-XIX-8 was drilled on the western slope of the tell, at an was divided into three stratigraphic units. elevation of 14.30 m above mean sea level (amsl) and 90 m from the Unit A, correspond to the substrate, located 860–750 cm below start of the ERT line Akko-1 (Figs. 3, 4 and 7). This site was selected to surface (180–70 cm below mean sea level; bmsl) in AK-XIX-6 and test the nature of the high resistivity features identified in the ERT 750–620 cm below surface (17 cm bmsl–113 cm amsl) in AK-XIX-7. It is profiles at >5m depth (Fig. 4). Two stratigraphic units are present in composed of fine to medium ochre-coloured quartz-rich sands contain­ core AK-XIX-8. ing cm-scale kurkar aggregates. This unit is characterised by negative Unit A (10–6 m depth), is composed of fine sands containing cm-scale magnetic susceptibility values (-10.10^-5 to -2.10^-5 SI for AK-XIX-6 and kurkar aggregates. MS values are close to zero and constant, character­ <-17.10^-5 SI for AK-XIX-7). istic of well sorted fine aeolian quartz-rich sands, as confirmed by Unit B, is composed of clean yellow sands between 750 and 300 cm observation under the stereomicroscope. below surface in AK-XIX-6 (70 cm bmsl–380 cm amsl) and between 600 Unit B (6–0 m depth) is dominated by brown silty-clay sediments and 405 cm below surface in AK-XIX-7 (113–328 amsl). MS values are with a subsidiary fraction of fine sands and some kurkar aggregates close to zero in AK-XIX-6 and slightly higher and less constant in AK-XIX- which is characteristic of an anthropogenic/agricultural soil. MS mea­ 7 due to the presence of sherds. In AK-XIX-7, the top of unit B is dated surements show highly variable values due to the presence of small from the Persian period and the sherds present sharp edges sign of an sherds and variable sediment texture. absence of wave action abrasion. Unit B is interpreted as a supralittoral Fig. 7. Core AK-XIX-8. This sediment core shows the presence of two distinct stratigraphic units. See Fig. 3 for location of the core. 73 M. Giaime et al. Quaternary International 602 (2021) 66–81 Fig. 8. Cores AK-XIX-6 and AK-XIX-7. The sediment cores show the presence of an supralittoral environment followed by the development of an agricultural soil. See Fig. 3 for location of the cores. Fig. 9. Sedimentary texture, magnetic susceptibility and analysis of ostracods in core AK-XIX-5. Three different environmental units were identified. The unit highlighted in light brown show the sewage/drainage identified in this core. Fauna is dominated by the ostracod specie Heterocypris salina. See Fig. 3 for location of the core. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) 74 M. Giaime et al. Quaternary International 602 (2021) 66–81 environment. amsl) correspond to the substrate. It is composed of fine ochre-coloured Unit C, correspond to the uppermost unit of both cores. It is domi­ quartz sands with cm-scale kurkar aggregates. MS shows slightly nega­ nated by brown silty-clay sediments with a subsidiary fraction of fine tive values (ca. -10.10^-5 SI). sands and some kurkar aggregates. MS values range from 0.10^-5 to Unit B, located from a depth of 720 to 600 cm (100–220 cm amsl), is 500.10^-5 SI due to lithological and textural heterogeneity and the composed of grey fine sediments (silts and fine sands) rich in organic presence of small sherds. It is interpreted as an anthropogenic/agricul­ matter. It is characterised by fluctuation in the MS values (from -4.10^-5 tural soil. SI to 470.10^-5 SI) due to changes in grain size throughout the unit and the presence of numerous ceramics (Table 3). Ostracod fauna is domi­ 4.3.3. Core AK-XIX-5 nated by the freshwater ostracod Heterocypris salina (>60% of the total Core AK-XIX-5 was drilled at an elevation of 8.19 m amsl (Figs. 3 and assemblage). This unit is radiocarbon dated to the Persian period 9). It can be divided into five different stratigraphic units. (750–420 cal BC at 678 cm depth (141 cm amsl)) and 770–520 cal BC at Unit A at the base of the core (900–720 cm depth; 81 cm bmsl–99 cm 710 cm depth (109 cm amsl). The chronology is confirmed by the Fig. 10. Sedimentary texture, magnetic susceptibility and analysis of ostracods of cores AK-XIX-3, AK-XIX-4, AK-XIX-9 and AK-XIX-10. The different environmental units identified shows the presence of a shallow coastal environment dated to the Persian period (units highlighted in blue). See Fig. 3 for location of the cores. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) 75 M. Giaime et al. Quaternary International 602 (2021) 66–81 presence of numerous ceramic sherds in the core from the same period anthropogenic infill. (Table 3). This unit containing a lot of anthropic material (charcoals, grape seeds, ceramic sherds) is interpreted as a possible sewage/ 4.3.4. Cores AK-XIX-3, AK-XIX-4, AK-XIX-9 and AK-XIX-10 drainage channel. These four cores were drilled further south (Fig. 3). As for the core Unit C is a 120 cm thick sandy layer (600–480 cm depth; 220–340 cm AK-XIX-5, their upper sections represent recent anthropogenic infills amsl) barren of fauna and characterised by fluctuating MS values. and post-habitation anthropogenic/agricultural soil. To ease visualiza­ Sherds found in this unit are dated from the Persian period. This unit tion, the upper sections of these cores are not represented in the strati­ may be inferred as a beach environment. graphic logs (Fig. 10). Unit D between 480 and 130 cm depth (340–790 cm amsl) is Unit A in the four cores correspond to the substrate, either composed dominated by brown silty-clay sediments with a subsidiary fraction of of fine quartz sands with centimetre scale kurkar aggregates (cores AK- fine sands, kurkar aggregates and small sherds. MS values fluctuate be­ XIX-3 and AK-XIX-4) or kurkar sandstone (cores AK-XIX-9 and AK-XIX- tween 0 and 150.10^-5 SI. This unit is interpreted as an anthropogenic/ 10). agricultural soil. Unit B, ranging from 50 to ca. 100 cm thick, is composed of inter­ Unit E, at the top of the core (130–0 cm depth), correspond to recent bedded layers of sand and small pebbles supported by a sandy matrix, Fig. 11. Sedimentary texture, magnetic susceptibility and analysis of ostracods of cores AK14-C5 and AK17-C7. The different environmental units identified are characteristic of a coastal progradation. The infralittoral environment in which we identified marine and coastal ostracods is followed by the development of a supralittoral beach environment barren of fauna. See Fig. 3 for location of the cores. 76 M. Giaime et al. Quaternary International 602 (2021) 66–81 both hosting shell fragments. MS values are relatively low, with some 5.1. Geomorphological setting of the western edge of Tel Akko spikes in the MS values, probably related to the presence of ceramic- sherds and or the presence of small rounded pebbles. The sediments Multidisciplinary investigation of the western edge of Tel Akko has contain shell fragments. Well-preserved shells are scarce with a few demonstrated that this part of the tell has been built on coastal sand marine (e.g. Donax trunculus, Conus mediterraneus, Nassarius corniculus) dunes covering the kurkar bedrock as shown by the analysis of the ERT and brackish water (Melanoides tuberculata, Melanopsis buccinoidea) profiles and coring. Core AK-XIX-8, drilled along the ERT line Akko-1 species identified. Ostracod fauna collected in the samples are repre­ (Fig. 3), allowed us to demonstrate that the lower resistivity areas in sentative of lagoonal (Cyprideis torosa), marine lagoonal (Xestoleberis the upper part of the ERT profiles correspond to brown silty-clay sedi­ spp.) and coastal/marine species. Among the coastal/marine groups ments with a subsidiary fraction of fine sands, kurkar aggregates and over 20 different species were identified (e.g. Aurila convexa, Lox­ small sherds, most probably related to anthropogenic/agricultural soils. oconcha, Pontocythere cf. elongata; Fig. 10). Even if the diversity of spe­ Stratigraphy of the core proved that the high resistivity areas below 6 m cies is high, the density is low with less than 100 ostracod valves per 10 g depth, reflect the presence of a sand unit. This sand is mainly composed of sediment except in core AK-XIX-3 where the density reaches of very fine quartz grains and reflect aeolian deposition of sand in the 500–1000 ostracod valves per 10 g of sediment. The upper part of this area. Archaeological evidence confirms the building of city ramparts unit is located above present mean sea level and may only show the throughout the tell using this sandy substratum (Fig. 4). deposition of marine infralittoral sediments in the mid-littoral zone. This Analysis of core AK-XIX-5, show the possible presence of a sewage/ unit is radiocarbon dated from the Early Iron Age (1010–900 cal BC) in drainage canal to the sea west of the substantial buildings identified and core AK-XIX-3 and from the Phoenician/Persian period (750–420 cal where ostraca were found in area G (Figs. 9 and 12). The presence of the BC) in core AK-XIX-4. Dating of core AK-XIX-4 is in agreement with the sewage is evidenced by a relatively fine sediment unit containing the chronology of all the four cores based on the ceramic sherds found ostracod Heterocypris salina which can inhabit brackish temporary water within this unit. However, the dating in AK-XIX-3 shows a chronological bodies (Fig. 9). It has also been demonstrated that H. salina can with­ gap likely due to the old wood effect of the charcoal (Table 2 and 3). Part stand a certain degree of organic water pollution (Mezquita et al., 1999). of the ceramic sherds found within this unit attest to weathering induced Furthermore, we found in this unit numerous pieces of charcoal, grapes in a high energy environment, associated with an ancient coastal surf seeds and sherds indicating the deposition of anthropogenic material zone (Figs. 2 Supplementary Material). originating from the tell during heavy rains in the Persian period Unit C is composed of fine yellow quartz-rich sand barren of fauna (Table 2). An even earlier sewage/runoff drainage was found in the and demonstrates the development of an emerging supralittoral envi­ southern limit of the tell, under area PH, dated to the MBII thought at ronment most probably associated with the progradation of the coast at first to be a postern gate (secondary door or gate in a fortification; that time. Raban, 1991), but should be identified as a sewage/drainage of the gate area of area P as it was too small to allow the passage of people (Fig. 13). 4.3.5. Cores AK14-C5 and AK17-C7 The results of the study of the cores drilled on the western and AK14-C5 and AK17-C7 were drilled in the southwest of the tell southwestern areas of Tel Akko outline the location the city’s waterfront (Figs. 3 and 11). Their general stratigraphy is composed of four main in the Phoenician/Persian period (Figs. 12 and 14). It was evidenced in units. the form of an exposed shallow sandy to pebbly coastal environment. Unit A corresponds to the kurkar substrate at 8.20 m depth (3.20 m Rounded sherds of imported ceramics found within the cores, show bmsl) in core AK14-C5 and 4.60cm depth (1.30 m bmsl) in core AK17- evidence of weathering by waves in the coastal surf zone and were dated C7. to the same period (Table 3; Supplementary Material). This unit directly Unit B is composed of fine marine quartz-rich sands characterised by covers the substrate composed of kurkar and fine to medium quartz-rich low MS values (close to 0.10^-5 SI). Ostracod fauna is scarce with density sands probably originating from the weathering of the kurkar sandstone <100 valves per 10 g of sediments. The faunal assemblage is charac­ and the deposition of aeolian sands. Considering its elevation in relation teristic of a coastal marine environment with lagoonal (Cyprideis torosa), to the current sea level and assuming that the relative sea level has been marine lagoonal (Xestoleberis spp.) and coastal/marine species (e.g. roughly stable since the end of the 1st Millennium BC (Sivan et al., Aurila convexa, Loxoconcha rubrincinata, Pontocythere cf. elongata). Two 2001), large ships linked to interregional trade could not dock in the freshwater species (Candona neglecta and Illyocypris gibba) were identi­ immediate proximity of the tell during the Persian period as the water fied in the lowermost part of the unit. A ceramic sherd dated to the depth would have been <50 cm (Figs. 10, 12 and 14). Moreover, we may Middle Bronze was found in core AK-14-C5 at 676 cm depth (175 cm reasonably expect a rapid infill of this area during the Late Persian bmsl). Neither marine shells nor organic material were found for period because the ceramic sherds discovered are dated from the 4th radiocarbon dating. This unit is interpreted as prograding beach century BC at the very latest. GPR surveys shows the presence of a ag­ environment. gradational and progradational geomorphic environment such as a Unit C is composed of quartz-rich sands located 40–60 cm amsl. It is coastline in the same area. characterised by low MS values typical of well sorted quartz sands. The fauna is absent from this unit. Unit B is interpreted as a emerging beach 5.2. Location of the anchorage environment. Unit D in the uppermost part of the core (first 320 cm in AK14C5 and The beach identified at the foot of Tel Akko provided direct access to first 120 cm in AK17-C7) corresponds either to an anthropogenic/agri­ the open sea but the shallow water depth did not allow for the naviga­ cultural soil of silty-clay to sandy texture or recent anthropogenic infills. tion of large vessels. Around our core AK14-C5 and further south, the area seems to have been more favourable for the anchorage of large 5. Discussion ships during the Phoenician and Persian periods (Fig. 12). Sediment cores corroborate the interpretation of the ERT profiles. While high re­ In the Iron Age – Phoenician/Persian periods – a dynamic beach sistivity units correspond to relatively coarse dry sands and kurkar, environment has been shown on the western edge of Tel Akko. At the lower resistivity sections (blue colour) are indicative of the presence of end of the Persian period, the increasing sedimentary inputs led to the fine to medium sands, typical of an open marine environment on the progressive infilling of the area and Tel Akko lost its access to the sea. southwest of Tel Akko (Figs. 5 and 14). Thus, the Phoenician/Persian anchorage, now landlocked, must have existed in the coastal plain located to that southwest area of the tell (Fig. 14). Persian ceramics were previously found in core AK-6 (Morhange et al., 2016, Fig. 14) at 575 cm 77 M. Giaime et al. Quaternary International 602 (2021) 66–81 Fig. 12. Correlation of sediment cores retrieved from the west and southwest feet of Tel Akko. Location of transects is given in Fig. 3. These transects show the geomorphology of the beach environment and the variation of the water column height. The different shades of blue depict the infralittoral zone (always submerged) from the mediolittoral zone (which corresponds to the transition zone between a permanently emerged and a permanently submerged environment). C14 dating of cores AK-XV-1 and AK-XV-2 (Giaime et al., 2018) have been calibrated using the new calibration curve IntCal20 (Reimer et al., 2020). See Fig. 3 for the location of the cores. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) depth (200 cm below mean sea level) showing that the water column in and Linder, 2004). This merchant boat, radiocarbon dated to about 400 the area neighbouring our core AK14-C5 was ca. 200 cm (Figs. 11 and years BC, had a draft of 140 cm and a cargo capacity of 15.9 tons (Ben 14). This depth was sufficient for the anchorage of maritime vessels Zeev et al., 2009). We can reasonably assume that this kind of boat was (Fig. 12). An example of the boats in use along the Levantine coast in the able to approach the coast of Tel Akko with a 200 cm water depth. Persian period is given by the Ma’agan Mikhael wreck, discovered ca. Moreover, goods could be safely unloaded and transported to the beach 40 km south of Tel Akko in 1985 (Linder and Kahanov, 2003; Kahanov by smaller ships such as lighters or barges. Another possibility that 78 M. Giaime et al. Quaternary International 602 (2021) 66–81 Fig. 13. Photo of the MBII (2000–1750 BC) postern gate in area P, from the south. Photo © M. Artzy. cannot be excluded, is the presence of basic constructions of a proto harbour, such as small sized piers, quays or jetties in the area investi­ gated which had facilitated the docking of boats. 5.3. Coastal progradation and abandonment of Tel Akko Radiocarbon dating and ceramic sherds identification confirms that the sea was present in the immediate proximity of the tell until the 3rd century BC. At that time, the sedimentary inputs on the western side of Tel Akko led to the progressive infilling of the area. The fact that Tel Akko benefited from a direct access to the sea from the Bronze Age to the Late Persian/Early Hellenistic period has largely contributed to its importance in maritime activities over the centuries in the Bay of Haifa/ Akko and at a larger regional scale (Artzy, 2006). The small apparent sedimentation rate in the area is in contradiction with the large sedi­ mentation rates evidenced in the centre of the Haifa Bay. Porat et al. (2008) have estimated the progradation rate of the bay to be ~40 cm/year between 4000 and 3650 years BP (ca. 2000–1650 BC). After­ ward, the coastline continued to migrate westwards at an approximate rate of 50 cm/year until present day. In the Akko plain, sediment deposition of that volume reported by Zviely et al. (2007) and Porat et al. (2008) is not apparent. We assume that an obstacle, in the form of an aeolianite ridge (kurkar) identified ca. 400 m to the south of the tell could have blocked the longshore drift and Fig. 14. Palaeoenvironmental reconstruction of Tel Akko maritime interface in the sedimentary inputs to the north (Fig. 14). This ridge probably acted the Phoenician/Persian period. The different environments have been identified as a natural coastal groyne or dyke. The relative influence of the river based on this study and from previous research (Morhange et al., 2016; Giaime and marine sediment flux by longshore drift in the loss of Tel Akko’s et al., 2018). The dotted line shows the location of the core transects as seen in direct access to the sea is still difficult to determine precisely. However, Fig. 12. Topography was adapted from a map produced by Trei­ del (1925–1926). our data point toward increasing sedimentary dynamics from marine origin in the southwestern area of Tel Akko leading to the progressive westward progradation of a sandy beach by continuous deposition of large-scale maritime activity at a harbour in the promontory prior to the coastal sediments during the Late Persian period. Hellenistic period (Stern et al., 2008). However, from that period, the Archaeological evidence attest that Tel Akko was occupied until the new city benefited by a commercial (Galili and Rosen, 2017; Galili et al., beginning of the Hellenistic period (Artzy, 2015). Archaeological find­ 2010) and a military harbour (Sharvit et al., 2013, 2021). Intense ings attest to extensive maritime trade reflecting the city’s growth and maritime activities since the Hellenistic period were evidences during prosperity until this period (Artzy and Be’eri, 2010). At the same time, the dredging of the harbour basin (Galili et al., 2010). These findings the population began relocating to the lower city on the Akko prom­ show that the new city of Akko became the main commercial centre at ontory. For a few decades, Tel Akko and the new city were occupied the expense of the tell. simultaneously (3rd-2nd centuries BC). In our work, we did not find evidence of any coastal or marine 6. Conclusion environment dated after the Late Persian/Early Hellenistic period. It seems that the main response to the increasing sedimentary input has Archaeological excavations have demonstrated that Tel Akko was been the relocation of the city to the Akko promontory in the Hellenistic densely occupied and was a main maritime hub along the coast of the period (Figs. 2 and 14). Rare archaeological finds dated prior to the 3rd southern Levant during the Phoenician/Persian periods. Intensive century BC and the complete absence of Attic (Greek) imported ware, maritime activities have been evidenced by the numerous imported commonly widespread in the coastal region, attest that there was no ceramics and ostraca discovered on the tell, suggesting that at that time, 79 M. Giaime et al. Quaternary International 602 (2021) 66–81 an anchorage was located in close proximity. Using a multidisciplinary Artzy, M., 2012. Return to Tel Akko, its Anchorages, Harbour, and Surroundings, vol. 37. Recanati Institute for Maritime Studies Newsletter, University of Haifa, pp. 5–14. approach to better understand the evolution of the Akko coastal plain to Artzy, M., 2015. What is in a name? ‘Akko–Ptolemais–‘Akka–Acre. Complutum 26 (1), the west and southwest of Tel Akko, we were able to locate the Phoe­ 205–212. nician/Persian waterfront at the southwestern foot of the tell. We sug­ Artzy, M., Be’eri, R., 2010. Tel Akko. In: Killebrew, A.E., Raz-Romero, V. (Eds.), One gest that this area was the most suitable for the anchorage of boats with a Thousand Nights and Days—Akko through the Ages. Hecht Museum, Haifa, pp. 14–23. water depth reaching ca. 2 m during the Persian period. We recommend Artzy, M., Quartermaine, J., 2014. How and when did Tel Akko get its unusual banana future research to be focused around this specific area of the tell. The shape? In: Galanakis, Y., Wilkinson, T., Bennet, J. (Eds.), Аθυρματα. Critical Essays new data findings highlight the archaeological potential of the coastal on the Archaeology of the Eastern Mediterranean in Honour of E. Susan Sherrat, 11- 22. Archaeopress Archaeology, UK. plain in immediate proximity to tell and further south-southwest. Local Artzy, M., Jol, H., Giaime, M., Salmon, Y., abu-Hamid, A., López, G.I., Morhange, C., and regional public authorities should be made aware of this new hy­ Kaniewski, D., Bauman, P., Killebrew, A.K., 2021. Coastline, river changes and their pothesis in the event of future urban developments planned in the area. effects on anchorages/harbors and habitation patterns: Akko (Israel) as an example. In: Demisticha, S., Blue, L. (Eds.), Under the Mediterranean I. Studies in Maritime Archaeology. Sidestone Press, Leiden, Netherlands, pp. 267–278. Author contributions Athersuch, J., Horne, D.J., Whittaker, J.E., 1989. Marine and brackish water ostracods (Superfamilies Cypridacea et Cytheracea): Keys et notes for the identification of the species. E.J. Brill, Leiden, The Netherlands. M.G., H.J., Y.S., G.L. and M.A. designed the study. All the authors Avnaim-Katav, S., Agnon, A., Sivan, D., Almogi-Labin, A., 2016. Calcareous assemblages collected the data in the field. M.G., L.B. and A.M. carried out laboratory of the southeastern Mediterranean low-tide estuaries. Seasonal dynamics and paleo- analyses. M.G., H.J., L.B. and M.A. analysed the resulting data. M.G., H. environmental implications. J. Sea Res. 108, 3049. Beeri, R., 2008. Tel Akko and the Urbanization of Akko Plain in the First Half of the J. and M.A. prepared the figures and wrote the original manuscript. All Second Millennium BCE. Unpublished PhD thesis. Department of Archaeology, the authors revised the original manuscript draft. University of Haifa (In Hebrew). Ben Zeev, A., Kahanov, Y., Tresman, J., Artzy, M., 2009. The Ma‘agan Mikhael Ship: A Reconstruction of the Hull. Israel Exploration Society and Leon Recanati Institute for Data availability Maritime Studies, University of Haifa, Haifa. Conyers, L.B., Goodman, D., 1997. Ground-penetrating Radar: an Introduction for Data used for the present research are detailed in the paper and its Archaeologists. AltaMira Press. D’Angelo, G., Garguillo, S., 1978. Guida Alle Conchiglie Mediterranee, Conoscerle, supplementary files. Cercarle, Collezionarle. Fabri Editori, Milan. Dothan, M., 1976. Akko: interim excavation report first season 1973/74. Bull. Am. Sch. Funding Orient. Res. 224, 1–48. Dothan, M., 1985a. A Phoenician inscription from Akko. Isr. Explor. J. 35 (2/3), 81–94. Dothan, M., 1985b. Ten seasons of excavations in ancient Akko. Qadmoniot 18, 2–24. This research has been carried out thanks to the support of the United Elyashiv, H., Bookman, R., Zviely, D., Avnaim-Katav, S., Sandler, A., Sivan, D., 2016. The States-Israel Binational Science Foundation (BSF) to the project “A interplay between relative sea-level rise and sediment supply at the distal part of the Nile littoral cell. Holocene 26 (2), 248–264. Palaeo-Hydrological Approach to Understanding the Complex Coastal Empereur, J.Y., 1995. Le port hellénistique d’Amathonte. In: Karageorghis, V., Evolution of an Ancient Maritime Trade City: The Tel Akko Multi- Michaelides, D.D. (Eds.), Cyprus and the Sea Proceedings of the International Disciplinary Case Study, Israel” to M. Artzy and H.M. Jol (Grant No. Symposium, 131-138. University of Cyprus-Cyprus Ports Authority: Nicosia. 2016080). Empereur, J.Y., Verlinder, C., 1987. The underwater excavations at the ancient port of Amathus in Cyprus. Int. J. Naut. Archaeol. 16, 7–18. Empereur, J.Y., Koželj, T., Picard, O., Wurch-Koželj, M., 2018. The Hellenistic Harbour of Declaration of competing interest Amathus: Underwater Excavations, 1984-1986. Volume 1, Architecture and History. Études Chypriotes (19). École Française d’Athènes, Athens. Frenzel, P., Boomer, I., 2005. The use of ostracods from marginal marine, brackish waters The authors declare that they have no known competing financial as bioindicators of modern and quaternary environmental change. Palaeogeogr. interests or personal relationships that could have appeared to influence Palaeoclimatol. Palaeoecol. 225 (1), 68–92. Galili, E., Rosen, B., 2017. The Akko marina archaeological project—summary. In: the work reported in this paper. Galili, E. (Ed.), The Akko Marina Archaeological Project. BAR Publishing, Oxford, England, pp. 320–344. BAR International Series 2862. Acknowledgements Galili, E., Rosen, B., Zviely, D., Silberstein, N., Finkielsztejn, G., 2010. The evolution of Akko harbour and its Mediterranean maritime trade links. J. I. Coast Archaeol. 5, 191–211. Geoarchaeological investigations at Tel Akko have been possible Gambash, G., 2014. En route to Egypt: Akko in the Persian period. J. Near E. Stud. 73 (2), thanks to permits provided by the Israel Antiquities Authority (IAA) and 273–282. Giaime, M., Morhange, C., Marriner, N., López, G.I., Artzy, M., 2018. Geoarchaeological financial support from the BSF (2017–2020). We thank Drs. Nimer Taha investigations at Akko, Israel: new insights into landscape changes and related and Nicolas D. Waldmann, laboratory manager and laboratory leader of anchorage locations since the Bronze Age. Geoarchaeology 33 (6), 641–660. the PetroLab, University of Haifa, for their analytical support and use of Giaime, M., Marriner, N., Morhange, C., 2019. Evolution of ancient harbours in deltaic their facilities. M.G. thanks the Authority for Advanced Studies and the contexts: a geoarchaeological typology. Earth Sci. Rev. 191, 141–167. Inman, D.L., 2003. Littoral cells. In: Schwartz, M. (Ed.), Encyclopedia of Coastal Science. Leon H. Charney School of Marine Sciences at the University of Haifa The Earth Sciences Encyclopedia Online. Kluwer Academic Publishers, Dordrecht, and the BSF for the funding of his post-doctoral fellowship. H.J. thanks Netherlands, pp. 594–599. the University of Wisconsin - Eau Claire’s Office of Research and Jol, H.M., Bristow, C.S., 2003. GPR in sediments: advice on data collection, basic processing and interpretation, a good practice guide. In: Bristow, C.S., Jol, H.M. Sponsored Programs for Student/Faculty Research, Collaboration and (Eds.), Ground Penetrating Radar in Sediments, 9-27. Geological Society, Special Diversity mentoring Grants for Logan Bergevin and Ethan Sailer- Publications, London, p. 211. Haugland. MA is thankful to the Hatter Foundation and an anonymous Kafri, U., Ecker, A., 1964. Neogene and Quaternary Subsurface Geology and Hydrogeology of the Zevulun Plain. Ministry of Development, Geological Survey: donor. We thank two anonymous reviewers for their careful reading of Israel. our manuscript and their insightful comments and suggestions. Kahanov, Y., Linder, E., 2004. In: Tresman, J. (Ed.), The Ma‘agan Ship: the Recovery of a 2400-Year-Old Merchant-Man, Volume II. Israel Exploration Society and University of Haifa, Jerusalem. Appendix A. Supplementary data Kaniewski, D., Van Campo, E., Morhange, C., Guiot, J., Zviely, D., Shaked, I., Otto, T., Artzy, M., 2013. Early urban impact on Mediterranean coastal environments. Sci. Supplementary data to this article can be found online at https://doi. Rep. 3, 3540. Lachenal, A.M., 1989. Écologie des ostracodes du domaine méditerranéen: Application org/10.1016/j.quaint.2021.06.008. au golfe de Gabès (Tunisie orientale), les variations du niveau marin depuis 30 000 ans, vol. 108. Documents des Laboratoires de Géologie de Lyon. References Linder, E., Raban, A., 1965. Underwater survey of Akko harbour. In: The Western Galilee and the Galilee Coast. The 19th Convention for Yediat Haaretz, Jerusalem, pp. 180–193 (In Hebrew). Artzy, M., 2006. The carmel coast during the second part of the late Bronze age: a center for eastern mediterranean transhipping. Bull. Am. Sch. Orient. Res. 343 (1), 45–64. 80 M. Giaime et al. Quaternary International 602 (2021) 66–81 Linder, E., Kahanov, Y., 2003. In: Black, E. (Ed.), The Ma‘agan Mikhael Ship: the Büntgen, U., Capano, M., Fahrni, S.M., Fogtmann-Schulz, A., Friedrich, R., Recovery of a 2400-Year-Old Merchant-Man, vol. I. Israel Exploration Society and Köhler, P., Kudsk, S., Miyake, F., Olsen, J., Reinig, F., Sakamoto, M., Sookdeo, A., University of Haifa, Jerusalem. Talamo, S., 2020. The IntCal20 Northern Hemisphere radiocarbon age calibration Marriner, N., Morhange, C., Kaniewski, D., Carayon, N., 2014. Ancient harbour curve (0-55 cal kBP). Radiocarbon 1–33. https://doi.org/10.1017/RDC.2020.41. infrastructure in the Levant: tracking the birth and rise of new forms of Sharvit, J., Buxton, B., Hale, J.R., Ratzlaff, A., 2021. The Hellenistic-Early Roman anthropogenic pressure. Sci. Rep. 4, 5554. Harbour of Akko: preliminary finds from archaeological excavations at the foor of Marriner, N., Morhange, C., Flaux, C., Carayon, N., 2017. Harbors and ports, ancient. In: the southeastern seawall at Akko, 2008-2014. In: Demisticha, S., Blue, L. (Eds.), Gilbert, A.S., Goldberg, P., Holliday, V.T., Mandel, R.D., Sternberg, R.S. (Eds.), Under the Mediterranean I. Studies in Maritime Archaeology. Sidestone Press, Encyclopedia of Geoarchaeology. Springer Netherlands, Dordrecht, pp. 382–403. Leiden, Netherlands, pp. 163–180. Mezquita, F., Griffiths, H., Sanz, S., Soria, J., Piñón, A., 1999. Ecology and distribution of Sharvit, J., Planer, D., Buxton, B., 2013. Preliminary findings from archaeological ostracods associated with flowing waters in the eastern iberian peninsula. J. Crustac excavations along the foot of the southern seawall of Akko, 2008–2012 Michmanim Biol. 19 (2), 344–354. 24, 39–40 (In Hebrew with English abstract). Morhange, C., Giaime, M., Marriner, N., abu Hamid, A., Bruneton, H., Honnorat, A., Sivan, D., Gvirtzman, G., Sass, E., 1999. Quaternary stratigraphy and paleogeography of Kaniewski, D., Magnin, F., Porotov, A.V., Wante, J., Zviely, D., Artzy, M., 2016. the Galilee coastal plain, Israel. Quat. Res. 51 (3), 280–294. Geoarchaeological evolution of Tel Akko’s ancient harbour (Israel). J. Archaeol. Sci.: Sivan, D., Wdowinski, S., Lambeck, K., Galili, E., Raban, A., 2001. Holocene sea-level Report 7, 71–81. changes along the Mediterranean coast of Israel, based on archaeological Pérès, J.M., 1982. Major benthic assemblages. In: Kinne, O. (Ed.), Marine Ecology, Part 1, observations and numerical model. Palaeogeogr. Palaeoclimatol. Palaeoecol. 167 vol. 5. Wiley, Chichester, pp. 373–522. (1–2), 101–117. Pérès, J.M., Picard, J., 1964. Nouveau Manuel de Bionomie Benthique de La Mer Stanley, J.D., Bernhardt, C.E., 2010. Alexandria’s Eastern Harbor, Egypt: pollen, Méditerranée. Periplus, Marseille. microscopic charcoal, and the transition from natural to human-modified basin. Porat, N., Sivan, D., Zviely, D., 2008. Late Holocene embayment infill and shoreline J. Coast Res. 26 (1), 67–79. migration, Haifa bay, eastern mediterranean. Isr. J. Earth Sci. 57 (1), 21–31. Stern, E., Galili, E., Rosen, B., 2008. Acco. In: Stern, E. (Ed.), The New Encyclopedia of Raban, A., 1983. The biblical port of Akko on Israel’s coast. Archaeology 36 (1), 60–61. Archaeological Excavations in the Holy Land, 24-36. Israel Exploration Society, Raban, A., 1985. The ancient harbours of Israel in Biblical times. Harbour Archaeology. Jerusalem. Biblical Archaeology Society: Washington, DC. In: Proceedings of the First International Workshop, vols. 11–44. BAR Publishing, Stuiver, M., Reimer, P.J., 1993. Extended 14C database and revised CALIB radiocarbon Oxford, England. calibration program. Radiocarbon 35, 215–230. Raban, A., 1991. The port city of Akko in the MBII. Michmanim 5, 17–34. Stuiver, M., Reimer, P.J., Reimer, R.W., 2020. CALIB 8.2 [WWW program] at. http:// Raban, A., 1993. A group of imported ‘east Greek’ pottery from locus 46 at area F on Tel calib.org. (Accessed 17 August 2020). Akko. In: Heltzer, M., Segal, A., Kaufman, D. (Eds.), Studies in Archaeology and Thompson, R., Oldfield, F., 1986. Environmental Magnetism. George Allen & Uniwin, History of Ancient Israel, in Honour of Moshe Dothan. University of Haifa, Haifa, London. pp. 73–98. Vachtman, D., Sandler, A., Greenbaum, N., Herut, B., 2013. Dynamics of suspended Raban, A., Stieglitz, R.R., 1993. Phoenicians on the Northern Coast of Israel in the sediment delivery to the Eastern Mediterranean continental shelf. Hydrol. Process. Biblical Period. Reuben and Edith Hecht Museum, University of Haifa, Haifa. 27, 1105–1116. Reimer, P.J., Austin, W.E.N., Bard, E., Bayliss, A., Blackwell, P.G., Bronk Ramsey, C., Zviely, D., Sivan, D., Ecker, A., Bakler, N., Rohrlich, V., Galili, E., Boaretto, E., Klein, M., Butzin, M., Cheng, H., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., Kit, E., 2006. Holocene evolution of the Haifa bay area, Israel, and its influence on Hajdas, I., Heaton, T.J., Hogg, A.G., Hughen, K.A., Kromer, B., Manning, S.W., ancient tell settlements. Holocene 16, 849–861. Muscheler, R., Palmer, J.G., Pearson, C., van der Plicht, J., Reimer, R.W., Zviely, D., Kit, E., Klein, M., 2007. Longshore sand transport estimates along the Richards, D.A., Scott, E.M., Southon, J.R., Turney, C.S.M., Wacker, L., Adolphi, F., Mediterranean coast of Israel in the Holocene. Mar. Geol. 237, 61–73. 81