Journal Pre-proof Using a multi-proxy approach to locate the elusive Phoenician/Persian anchorage of Tel Akko (Israel) Matthieu Giaime, Harry M. Jol, Yossi Salmon, Gloria I. López, Amani abu Hamid, Logan Bergevin, Paul Bauman, Alastair McClymont, Ethan Sailer-Haugland, Michal Artzy PII: S1040-6182(21)00334-7 DOI: https://doi.org/10.1016/j.quaint.2021.06.008 Reference: JQI 8915 To appear in: Quaternary International Received Date: 31 August 2020 Revised Date: 31 May 2021 Accepted Date: 6 June 2021 Please cite this article as: Giaime, M., Jol, H.M., Salmon, Y., López, G.I., Hamid, A.a., Bergevin, L., Bauman, P., McClymont, A., Sailer-Haugland, E., Artzy, M., Using a multi-proxy approach to locate the elusive Phoenician/Persian anchorage of Tel Akko (Israel), Quaternary International (2021), doi: https:// doi.org/10.1016/j.quaint.2021.06.008. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2021 Published by Elsevier Ltd. 1 Using a multi-proxy approach to locate the elusive Phoenician/Persian anchorage of 2 Tel Akko (Israel) 3 4 Matthieu Giaimea,b,*, Harry M. Jolc, Yossi Salmona, Gloria I. Lópeza, Amani abu Hamida,d, Logan Bergevinc, 5 Paul Baumane, Alastair McClymonte, Ethan Sailer-Hauglandc, Michal Artzya 6 a Hatter Laboratory, Recanati Institute for Maritime Studies, Department of Maritime Civilizations, of 7 ro 8 University of Haifa, Abba Khoushy Ave 199, Haifa, 3498838, Israel b 9 Impactant, Institut de Ciència i Tecnologia Ambientals (ICTA-UAB), Universitat Autònoma de Barcelona, -p 10 08193 Cerdanyola del Vallès, Barcelona, Spain re c 11 Department of Geography & Anthropology, University of Wisconsin-Eau Claire, 105 Garfield Avenue, lP 12 Eau Claire, WI 54701 USA na d 13 Israel Antiquities Authority, Rockefeller Archeological Museum, P.O.B 586, Jerusalem 9100402, Israel ur e 14 BGC ENGINEERING INC. Suite 500 - 1000 Centre St. NE, Calgary, AB, T2E 7W6, Canada Jo 15 * Corresponding author. 16 E-mail address: matthieu.giaime@gmail.com (M. Giaime). 17 18 19 20 21 22 23 24 25 Abstract 26 Previous geoarchaeological research on the Akko coastal plain have contributed to the 27 understanding of the ancient coastal interface and added evidence as to the location/shift of the 28 ancient anchorages dating from the Middle Bronze Age (beginning of the 2nd Millennium BC) to 29 the Early Hellenistic period (mid of the 2nd century BC) of the ancient site of Tel Akko. The present 30 research provides new insights into the environmental changes and likely anchorage sites along 31 the western edge of Tel Akko in the 1st Millennium BC (Iron Age II and III, periods associated with of 32 the Phoenician mariners and Persian army incursion). Our approach for locating the anchorage is ro 33 based on a detailed investigation of subsurface sediments combining sedimentological and faunal 34 -p analysis and radiocarbon dating of cores as well as identification of ceramic sherds found in the re 35 cores, and ground penetrating radar (GPR) and electrical resistivity tomography (ERT) surveys. lP 36 Paleoenvironmental changes are compared and contrasted with the results of the archaeological 37 investigations on the tell and in its vicinity. Our new data demonstrates that the na 38 Phoenician/Persian maritime interface of Tel Akko was mainly oriented toward the southwestern ur 39 area of the tell where a natural anchorage was likely to have been located. At that time, the water Jo 40 depth in this area was ca. 2m, allowing for the anchorage of seagoing vessels. Increasing sediment 41 deposition lead to the deterioration of direct, and eventual loss of access, to the sea. These 42 conditions initiated the abandonment of the tell in the Early Hellenistic period as well as the 43 westward shift to habitation on the peninsula, now the ‘Old city of Akko’, the Crusaders’ Saint 44 Jean d’Acre. 45 46 Keywords: Geoarchaeology; Geomorphology; Anchorages/harbors; Bio-sedimentology; Ground 47 penetrating radar (GPR), Electrical resistivity tomography (ERT) 48 49 50 1. Introduction 51 The layout of ancient coastal cities is dictated by the shoreline position as well as any neighbouring river. 52 In dynamic geomorphic settings, palaeoenvironmental reconstruction of ancient coastal settlements are 53 particularly challenging unless a multi-disciplinary approach is taken. The natural anchorages or artificial 54 harbours – lifelines of ancient coastal sites – may evolve or be relocated over time. This depends on the 55 change in orientation of the coastline due to sea-level changes, sedimentary inputs, coastal erosion and of 56 associated hazards (Giaime et al., 2019). In the 3rd and 2nd Millennia BC, the eastern Mediterranean ro 57 populations favoured natural coastal features as anchorage areas whereas artificial harbour basins and 58 -p related port facilities started to be developed mainly during the Iron Age (Marriner et al., 2014; Marriner re 59 et al., 2017). Along the coast of the southern Levant, the lack of bays on the coast, encouraged the lP 60 utilizations of river-mouths for maritime activities during the Bronze Age (Raban, 1985; Figure 1A). River- 61 mouths have generally been used as safe havens for relatively short time periods. Over the long term na 62 they were vulnerable to a multitude of natural hazards, which include infilling by both fluvial and marine ur 63 sediments due to sea level stabilisation (Giaime et al., 2019). Jo 64 The UNESCO World Heritage site of Akko/Acre is one of the oldest continuously inhabited sites 65 along the Levantine coast although its habitation pattern has evolved through time (general chronology 66 given in Table 1). The main change occurred at the beginning of the Hellenistic period when the tell was 67 deserted in favour of the Akko promontory where the city is presently situated (Artzy, 2015; Figure 2). 68 69 The first archaeological remains at Tel Akko are dated to the beginning of the Early Bronze Age, 70 ca. 5000 years ago. The first phase of urbanisation took place on the tell at the beginning of the Middle 71 Bronze. An impressive rampart reaching a height of 22 m and a width of 60 m dated to the Middle 72 Bronze IIa period was erected on the northern side of the tell. The highest part of it was located in Area 73 AB, where a large MBII building was constructed (Dothan, 1976; Raban, 1983, 1991; Be’eri 2008; Figure 74 3). The development of this urban centre deeply transformed the pre-existing forest ecosystems toward 75 an urban environment with a vegetation biome dominated by cultivated species and urban weeds 76 (Kaniewski et al., 2013). Written sources, including letters from the rulers of Akko, found mainly in 77 Egypt, identifies the city as an important harbour town serving the Egyptian authorities in the mid- 78 second Millennium BC (Artzy, 2012), although excavations have only located the latest part of the Late 79 Bronze (late second Millennium BC) until now. of 80 The second period of economic growth at Tel Akko is centred on the second part of the Iron Age ro 81 including the Persian Period at which time, signs of iron works were noted. The habitation pattern and 82 -p the population density increased after ca. 700-600 years BC (Artzy and Be’eri, 2010). The importance of re 83 Tel Akko in the trading network of the Levantine coast during the Iron Age is attested by the presence of lP 84 Phoenicians, Persians and Greeks merchants (Gambash, 2014). It wasn’t until the Hellenistic period (3rd 85 and early 2nd century BC) that the centre of the town moved to the promontory, where the ‘Old City of na 86 Akko/Acre is located (Artzy, 2015; Figure 2). ur 87 By using a multidisciplinary geo-environmental approach, we aim to present the Jo 88 geomorphological setting of the western area of Tel Akko in the Iron Age where a marine environment 89 has been evidenced in our previous research (Morhange et al., 2016; Giaime et al., 2018). We seek to 90 determine the possibilities of an anchorage and associated environmental limitations for that area. The 91 reconstruction of environmental changes affecting the western area of Tel Akko is fundamental for the 92 understanding of the development of the tell as well as the evolution of the habitation patterns during 93 the Phoenician/Persian period. 94 2. In search of Tel Akko’s Phoenician/Persian anchorage: (geo)archaeological context 95 The coastal site of Tel Akko was founded on the northern limit of the Haifa/Akko Bay atop a 96 naturally-occurring sand and kurkar feature. At Tel Akko, the kurkar, a local eolianite, is part of the 97 Kurdane formation (Kafri and Ecker, 1964) which is pre-Pleistocene of age (Sivan et al., 1999). The site is 98 located in a dynamic geomorphological environment at the end of the Nile’s littoral cell and in the 99 proximity of the Na’aman River (Belos) and its mouth (Morhange et al., 2016; Giaime et al., 2018; Figure 100 1.B). Haifa/Akko bay was formed by the subsidence of a geological graben delimited by two normal of 101 faults (Figure 1.B). It has undergone significant landscape changes during the Holocene (Zviely et al., ro 102 2006; Elyashiv et al., 2016; Figure 1.B). Following postglacial sea-level rise and subsequent flooding of 103 -p the graben, the shoreline transgressed several kilometres to the east, reaching its maximum landward re 104 position ca. 4000 years ago (Zviely et al., 2006). During the past 4000 years, the coastline rapidly lP 105 prograded to the west. The main driver of this progradation was the deposition of 700 million m3 of 106 Nile-derived quartz-rich sands transported northerly by longshore currents (Zviely et al., 2007) while na 107 fluvial sediment inputs from the local rivers appears as a secondary factor with exceptions for the ur 108 northern and southern extremities of the bay where the two main coastal streams flow into the sea Jo 109 (Vachtman et al., 2013; Zviely et al., 2006). 110 An artificial ‘Phoenician-style’ harbour was identified following submarine archaeological 111 prospections around the ‘Tower of flies’ which is still a landmark feature in the Bay (Linder and Raban, 112 1965; Raban and Stieglitz, 1993; Figure 2). It has been dated to the 6th century BC, the time of the of the 113 Persian king, Cambysis’ incursion during his conquest of Egypt. However, more recent studies have 114 qualified this dating. Artzy (2012) proposes that ‘Phoenician-type’ harbours continued to be 115 constructed in the south-eastern Mediterranean well after the Phoenician Period. One example is the 116 harbour at Amathus, Cyprus (Empereur and Verlinder, 1987; Empereur, 1995; Empereur et al., 2018) 117 which, despite depicting some Hellenic elements, was still constructed mainly in the Phoenician style 118 around 300 BC. In addition, a re-study of the ceramics from Raban’s excavations, have not shown any 119 pre-Hellenistic ceramic evidence. In addition, archaeological excavations in the Akko promontory where 120 the ‘Old City’ is located have not shed light on pre-Hellenistic occupation. Moreover, dredging of Akko’s 121 fishermen’s harbour has not provided evidence for the presence of pre-Hellenistic material in the 122 harbour basin (Galili et al., 2010; Galili and Rosen, 2017). Gambash (2014) states that during the the mid 123 6th century BC of the Persian period, Akko did not serve as a logistical centre and regular point d’appui 124 for Persian armies and fleets. It was only in the late 5th and 4th centuries BC that the Persian gathered of 125 their army of mercenaries at Tel Akko under Artaxerses II and III (Gambash, 2014; Diodorus Siculus ro 126 XV.41). 127 -p On Tel Akko, archaeological finds support the fact that the tell was the main centre of the area re 128 during the Persian period. On the summit of the tell (Area A, Figure 3), a Phoenician ostracon (a lP 129 potsherd used as a writing surface), the longest ever found in Phoenicia, was unearthed (Dothan, 130 1985a). More Phoenician ostraca, not yet published, were found in Area G, on the western side of the na 131 tell (Figure 3), where a massive building was noted as well (Artzy and Quatermaine, 2014). The latest ur 132 coins found in Area G are dated to the first half of the second century BC (Artzy, 2012). Rich remains Jo 133 from the Late Persian Period (late 5th-4th century BC) as well as numerous imports from the Aegean 134 world were found during excavations (Artzy and Be’eri, 2010; Dothan, 1976; 1985b), especially in Area F 135 in the north-western side of the tell (Raban, 1993; Figure 3). Area T, also on the western side of the tell, 136 was studied during a salvage excavation in 2010 directed by Abu Hamid of the Israel Antiquities 137 Authority (IAA). Results showed the presence of a small industrial area dated to the 4th century BC, 138 where Late Persian to Early Hellenistic ceramics were found (Artzy, 2012; Figure 3), as well as a gravelly- 139 sand facies bearing shells and ceramics interpreted as a high-energy coastline (Artzy, 2012). Pit surveys 140 carried out, as part of the ‘Total Archaeology’ Project, directed by Killebrew and Artzy since 2010, 141 accentuate the extent of 5th-4th century BC finds including an unusual large numbers of sherds, 142 especially of Maritime Transport Containers, both local and imported. 143 Based on the above evidence, Tel Akko was a major commercial hub of the Mediterranean trade 144 network in the Persian period, so it is hard to assume that its harbour was situated in the bay, 1.5 km 145 from the main habitation and storage area. 146 147 3. Material and methods of 148 Over the past ten years, geoarchaeological studies have shed light on the importance of fluvio-coastal ro 149 changes and possible location of Tel Akko’s anchorage(s) as early as the Middle Bronze Age (Morhange 150 -p et al., 2016; Giaime et al., 2018; Artzy et al., 2021). The focus of this paper lays on the geomorphological re 151 evolution and the possibilities of anchorage on the western edge of the tell (Figure 2 and 3). We carried lP 152 out a detailed sub-surface multi-disciplinary investigation. Ground penetrating radar (GPR) and electrical 153 resistivity tomography (ERT) surveys combined with sedimentological and faunal analysis of sedimentary na 154 cores. Cores were radiocarbon dated and ceramic sherds found in the cores were analysed to provide ur 155 high-precision relative chronology. Sediment cores, GPR grids and ERT lines have been geographically Jo 156 located with a Differential Global Positioning System (DGPS) and altitude benchmarked relative to 157 present mean sea level (msl) with a Total Station. 158 159 3.1. Electrical Resistivity Tomography (ERT) 160 The ERT profiles were collected in two phases: ERT profiles AKKO1, AKKO2, and AKKO3 were acquired in 161 July, 2011, and ERT profiles AKKO14, AKKO16, and AKKO17 were acquired in February, 2018. The ERT 162 profiles collected in 2011 were designed to investigate targets on the flanks of Tel Akko to depths of up 163 to 20 m below ground surface (mbgs), and were acquired using an ABEM SAS 3000 resistivity system, a 164 minimum electrode spacing of 1.5 m, and with standard Wenner array acquisition sequences. The 2018 165 profiles were designed to investigate deeper geological layers surrounding Tel Akko and were acquired 166 with a multichannel ABEM Terrameter LS unit, a minimum electrode spacing of 5.0 m, and expanded 167 gradient array acquisition sequences. Positions and topographic profiles for each ERT line were surveyed 168 using handheld Trimble Geo XH GPS units, with a nominal accuracy in vertical and horizontal directions 169 of less than 1 m. 170 For the 2011 and 2018 ERT surveys, resistivity data were collected through a linear array of electrodes 171 coupled to a direct current (DC) resistivity transmitter and receiver, and an electronic switching box. The of 172 spacing between electrodes largely controls the horizontal and vertical resolution of the data (smaller ro 173 spacing results in higher horizontal and vertical resolution). Similarly, the length of the array controls the 174 -p depth of investigation (longer arrays yield greater investigation depths). Data collection was carried out re 175 in a sequential and automated fashion that takes advantage of all possible combinations of current lP 176 injection and potential measurement electrodes. The data are downloaded to a computer for processing 177 and analysis. The data were then inverted using a two-dimensional (2-D) finite difference or finite na 178 element inversion routine using the RES2DINV software. To enhance layer boundaries at shallow depths, ur 179 the 2011 ERT profiles were inverted using a robust (L1-norm) constraint on the model. Conversely, due Jo 180 to lower resistivity contrasts between subsurface layers surrounding Tel Akko, a smooth model (L2- 181 norm) constraint was used for the 2018 ERT profiles. The model output from each ERT line is a 2-D cross- 182 section plotting resistivity (in ohm-m), versus depth. All ERT sections are presented as colour grids with 183 warm colours (reds and pinks) indicating relatively high resistivities and cool colours (light and dark 184 blues) showing regions of low resistivities. 185 186 3.2. Ground penetrating radar (GPR) 187 GPR can be used to investigate and identify human-made and/or geomorphic features based on the 188 sediments’ dielectric permittivity and electrical conductivity (Conyers and Goodman, 1997). GPR is a 189 non-invasive near-surface geophysical surveying technique that uses electromagnetic waves to image 190 the subsurface (Jol and Bristow, 2003). 191 A GPR survey was collected using a Sensors & Software pulseEKKO 1000 GPR system with an 192 antennae frequency of 225 MHz (0.1 m step size and 0.5 m antenna separation). The grid dimensions 193 were 24 m x 10.75 m (individual lines are 0.25 m apart). Site determination was based on archaeological 194 excavations (Td) on the tell. 195 After grid collection occurred, GPR lines were imported into GFP_Edit 4 (GPR Files and of 196 Parameters) and an axis-based grid was created. The created GFP file is then exported to a Sensors and ro 197 Software GPR processing application - EKKO_Project 5. GPR profiles were processed in a software 198 -p extension called SliceView (gain adjusted between 7 to 8). Conductive sediments affected the average re 199 velocity for the grid and was calculated to be between .07 and .08 m/ns based on hyperbolic lP 200 measurements. 201 na 202 3.3. Collection and analysis of the cores ur 203 Ten sediment cores were collected in plastic liners using a hydraulic percussion driller (Geoprobe DT22). Jo 204 Prior to split-opening and logging the sediment cores, every core section was initially run through a 205 Multi-Sensor Core Logger (MSCL) (PetroLab facility, University of Haifa). Magnetic susceptibility (MS) 206 measurements were key in deciphering between “natural” sedimentation (i.e., weak or negative MS 207 values produced by diamagnetic properties of well sorted quartz sand layers; Thompson and Oldfield, 208 1986) and more anthropogenic-like signatures (characterised by fluctuating MS values mainly due to the 209 presence of ceramics or changes in sediment texture). Core sub-samples from every stratigraphic unit 210 were taken for micropaleontological and sedimentological analyses. General sediment texture, i.e., 211 gravel (> 2 mm), sand (63 μm–2 mm) and silty-clay fractions (<63 μm), was determined by wet-sieving. 212 Ostracods were picked from the sand fraction >125 μm and identified using atlases and specific 213 literature (Athersuch et al., 1989; Lachenal, 2000; Frenzel and Boomer, 2005; Avnaim-Katav et al., 2016). 214 Molluscs were identified only from the gravel fraction (D’Angelo and Garguilo, 1978; Avnaim-Katav et 215 al., 2016). 216 217 3.4. Chronology 218 Four samples (grape seeds and charcoals; Table 2) were radiocarbon dated by Accelerator Mass 219 Spectrometry (AMS) at the Radiochronology Laboratory of the Centre d’études nordiques (University of of 220 Laval, Canada) and the Keck Carbon Cycle AMS Facility (University of California–Irvine, USA). ro 221 Identification of several ceramic fragments recovered from the cores provides a high-precision relative 222 -p chronology for specific stratigraphic units (Table 3; Supplementary Material 1). re 223 lP 224 4. Results 225 4.1. ERT measurements na 226 ERT lines Akko-1, Akko-2 and Akko-3 were surveyed on the western slope of the tell (Figure 3 and 4). ur 227 They demonstrate the presence of a relatively low to moderate resistive material (blue to green colours Jo 228 in the image) in the first 4-6 m below surface and interpreted as fine-grained sediments such as an 229 anthropogenic/agricultural soil. Highly resistive features (>200 ohm-m; red to pink colours), interpreted 230 as a kurkar outcrop or sandy unit (both exhibit the same high resistivity values in the ERT profiles) are 231 shown deeper in the subsurface. 232 Three 450-500 m-long ERT lines were surveyed in order to map the subsurface sedimentary 233 sequence of the coastal plain. Lines Akko-14 and Akko-16 have a mostly north/south orientation, 234 whereas Akko-17 trends approximately west/east (Figure 3 and 5). Most of the sections are dominated 235 by a low resistivity (blue colour) material interpreted as fine-grained (fine sands and silty-clays) over- 236 saturated sediments. Discrete, moderately resistive features are present at 10-20m depth. Only the 237 northern part of ERT lines Akko-14 and Akko-16 shows high resistive material (pink colour) interpreted 238 to be dry soil and colluvium from the tell, overlying kurkar sandstone (Figure 5) although it is challenging 239 to differentiate the nature of the subsurface sediments in this area as kurkar, dry sands and colluvium 240 will exhibit the same high resistivity values. 241 242 4.2. GPR 243 Ground penetrating radar data was collected on the western portion of the tell (near Td excavation) of 244 where a coastal environment was interpreted based on archaeological finds. The upper planview image ro 245 is a depth slice from between 0.44 m to 0.55 m depth. A curvilinear feature is highlighted in this depth 246 -p horizon. The lower image is individual Line 9 of the grid data. From 0 m to 12 m, multiple subhorizontal re 247 to horizontal continuous reflections become dipping reflections from 12 m to 20 m. The reflection lP 248 pattens discovered within the GPR data collected are interpreted as an aggradational and 249 progradational geomorphic environment such as a coastline. na 250 ur 251 4.3. Core analyses Jo 252 Stratigraphic units were described for each of the 10 cores analysed for this study according to 253 sedimentary texture and colour, MS properties and eventually faunal content. The environment 254 identified for each stratigraphic unit is interpreted according to the Mediterranean benthic classification 255 system (Pérès and Picard, 1964; Pérès, 1982; Supplementary Material 2). Analysis of the ceramic sherds 256 found in the cores confirms the robustness of the radiocarbon chronology measurements, however, 257 radiocarbon dating of one piece of charcoal shows a slight discrepancy that could be attributed to the 258 old wood effect or the reworking of older material (Sample ULA-8700; Table 2). 259 260 4.3.1. Core AK-XIX-8 261 Core AK-XIX-8 was drilled on the western slope of the tell, at an elevation of 1430 cm above mean sea 262 level (amsl) and 90 m from the start of the ERT line Akko-1 (Figures 3, 4 and 7). This site was selected to 263 test the nature of the high resistivity features identified in the ERT profiles at >5m depth (Figure 4). Two 264 stratigraphic units are present in core AK-XIX-8. 265 Unit A (10 to 6 m depth), is composed of fine sands containing cm-scale kurkar aggregates. MS 266 values are close to zero and constant, characteristic of well sorted fine aeolian quartz-rich sands, as 267 confirmed by observation under the stereomicroscope. of 268 Unit B (6 to 0 m depth) is dominated by brown silty-clay sediments with a subsidiary fraction of ro 269 fine sands and some kurkar aggregates which is characteristic of an anthropogenic/agricultural soil. MS 270 -p measurements show highly variable values due to the presence of small sherds and variable sediment re 271 texture. lP 272 273 4.3.2. Cores AK-XIX-6 and AK-XIX-7 na 274 AK-XIX-6, drilled at an elevation of 680 cm amsl and AK-XIX-7 (733 cm amsl) are the northernmost cores ur 275 of this study (Figure 3 and 8). The core was divided into three stratigraphic units. Jo 276 Unit A, correspond to the substrate, located 860-750 cm below surface (180-70 cm below mean 277 sea level; bmsl) in AK-XIX-6 and 750-620 cm below surface (17 cm bmsl- 113 cm amsl) in AK-XIX-7. It is 278 composed of fine to medium ochre-coloured quartz-rich sands containing cm-scale kurkar aggregates. 279 This unit is characterised by negative magnetic susceptibility values (-10.10^-5 to -2.10^-5 SI for AK-XIX-6 280 and <-17.10^-5 SI for AK-XIX-7). 281 Unit B, is composed of clean yellow sands between 750 and 300 cm below surface in AK-XIX-6 282 (70 cm bmsl-380 cm amsl) and between 600 and 405 cm below surface in AK-XIX-7 (113-328 amsl). MS 283 values are close to zero in AK-XIX-6 and slightly higher and less constant in AK-XIX-7 due to the presence 284 of sherds. In AK-XIX-7, the top of unit B is dated from the Persian period and the sherds present sharp 285 edges sign of an absence of wave action abrasion. Unit B is interpreted as a supralittoral environment. 286 Unit C, correspond to the uppermost unit of both cores. It is dominated by brown silty-clay 287 sediments with a subsidiary fraction of fine sands and some kurkar aggregates. MS values range from 288 0.10^-5 to 500.10^-5 SI due to lithological and textural heterogeneity and the presence of small sherds. 289 It is interpreted as an anthropogenic/agricultural soil. 290 of 291 4.3.3. Core AK-XIX-5 ro 292 Core AK-XIX-5 was drilled at an elevation of 819 cm amsl (Figures 3 and 9). It can be divided into five 293 different stratigraphic units. -p re 294 Unit A at the base of the core (900-720 cm depth; 81 cm bmsl- 99 cm amsl) correspond to the lP 295 substrate. It is composed of fine ochre-coloured quartz sands with cm-scale kurkar aggregates. MS 296 shows slightly negative values (ca. -10.10^-5 SI). na 297 Unit B, located from a depth of 720 to 600 cm (100-220 cm amsl), is composed of grey fine ur 298 sediments (silts and fine sands) rich in organic matter. It is characterised by fluctuation in the MS values Jo 299 (from -4.10^-5 SI to 470.10^-5 SI) due to changes in grain size throughout the unit and the presence of 300 numerous ceramics (Table 3). Ostracod fauna is dominated by the freshwater ostracod Heterocypris 301 salina (>60% of the total assemblage). This unit is radiocarbon dated to the Persian period (750-420 cal 302 BC at 678 cm depth (141 cm amsl)) and 770-520 cal BC at 710 cm depth (109 cm amsl). The chronology 303 is confirmed by the presence of numerous ceramic sherds in the core from the same period (Table 3). 304 This unit containing a lot of anthropic material (charcoals, grape seeds, ceramic sherds) is interpreted as 305 a possible sewage/drainage channel. 306 Unit C is a 120 cm thick sandy layer (600-480 cm depth; 220-340 cm amsl) barren of fauna and 307 characterised by fluctuating MS values. Sherds found in this unit are dated from the Persian period. This 308 unit may be inferred as a beach environment. 309 Unit D between 480-130 cm depth (340-790 cm amsl) is dominated by brown silty-clay 310 sediments with a subsidiary fraction of fine sands, kurkar aggregates and small sherds. MS values 311 fluctuate between 0 and 150.10^-5 SI. This unit is interpreted as an anthropogenic/agricultural soil. 312 Unit E, at the top of the core (130-0 cm depth), correspond to recent anthropogenic infill. of 313 ro 314 4.3.4. Cores AK-XIX-3, AK-XIX-4, AK-XIX-9 and AK-XIX-10 315 -p These four cores were drilled further south (Figure 3). As for the core AK-XIX-5, their upper sections re 316 represent recent anthropogenic infills and post-habitation anthropogenic/agricultural soil. To ease lP 317 visualization, the upper sections of these cores are not represented in the stratigraphic logs (Figure 10). 318 Unit A in the four cores correspond to the substrate, either composed of fine quartz sands with na 319 centimetre scale kurkar aggregates (cores AK-XIX-3 and AK-XIX-4) or kurkar sandstone (cores AK-XIX-9 ur 320 and AK-XIX-10). Jo 321 Unit B, ranging from 50 to ca. 100 cm thick, is composed of interbedded layers of sand and small 322 pebbles supported by a sandy matrix, both hosting shell fragments. MS values are relatively low, with 323 some spikes in the MS values, probably related to the presence of ceramic-sherds and or the presence 324 of small rounded pebbles. The sediments contain shell fragments. Well-preserved shells are scarce with 325 a few marine (e.g. Donax trunculus, Conus mediterraneus, Nassarius corniculus) and brackish water 326 (Melanoides tuberculata, Melanopsis buccinoidea) species identified. Ostracod fauna collected in the 327 samples are representative of lagoonal (Cyprideis torosa), marine lagoonal (Xestoleberis spp.) and 328 coastal/marine species. Among the coastal/marine groups over 20 different species were identified (e.g. 329 Aurila convexa, Loxoconcha, Pontocythere cf. elongata; Figure 10). Even if the diversity of species is high, 330 the density is low with less than 100 ostracod valves per 10 grams of sediment except in core AK-XIX-3 331 where the density reaches 500-1000 ostracod valves per 10grams of sediment. The upper part of this 332 unit is located above present mean sea level and may only show the deposition of marine infralittoral 333 sediments in the mid-littoral zone. This unit is radiocarbon dated from the Early Iron Age (1010-900 cal 334 BC) in core AK-XIX-3 and from the Phoenician/Persian period (750-420 cal BC) in core AK-XIX-4. Dating of 335 core AK-XIX-4 is in agreement with the chronology of all the four cores based on the ceramic sherds 336 found within this unit. However, the dating in AK-XIX-3 shows a chronological gap likely due to the old of 337 wood effect of the charcoal (Tables 2 and 3). Part of the ceramic sherds found within this unit attest to ro 338 weathering induced in a high energy environment, associated with an ancient coastal surf zone 339 (Supplementary Material 3). -p re 340 Unit C is composed of fine yellow quartz-rich sand barren of fauna and demonstrates the lP 341 development of an emerging supralittoral environment most probably associated with the progradation 342 of the coast at that time. na 343 ur 344 4.3.5. Cores AK14-C5 and AK17-C7 Jo 345 AK14-C5 and AK17-C7 were drilled in the southwest of the tell (Figure 3 and 11). Their general 346 stratigraphy is composed of four main units. 347 Unit A corresponds to the kurkar substrate at 820 cm depth (320 cm bmsl) in core AK14-C5 and 348 460 cm depth (130 cm bmsl) in core AK17-C7. 349 Unit B is composed of fine marine quartz-rich sands characterised by low MS values (close to 350 0.10^-5 SI). Ostracod fauna is scarce with density <100 valves per 10 grams of sediments. The faunal 351 assemblage is characteristic of a coastal marine environment with lagoonal (Cyprideis torosa), marine 352 lagoonal (Xestoleberis spp.) and coastal/marine species (e.g. Aurila convexa, Loxoconcha rubrincinata, 353 Pontocythere cf. elongata). Two freshwater species (Candona neglecta and Illyocypris gibba) were 354 identified in the lowermost part of the unit. A ceramic sherd dated to the Middle Bronze was found in 355 core AK-14-C5 at 676 cm depth (175 cm below present mean sea level). Neither marine shells nor 356 organic material were found for radiocarbon dating. This unit is interpreted as prograding beach 357 environment. 358 Unit C is composed of quartz-rich sands located 40-60 cm amsl. It is characterised by low MS 359 values typical of well sorted quartz sands. The fauna is absent from this unit. Unit B is interpreted as a 360 emerging beach environment. of 361 Unit D in the uppermost part of the core (first 320 cm in AK14C5 and first 120 cm in AK17-C7) ro 362 corresponds either to an anthropogenic/agricultural soil of silty-clay to sandy texture or recent 363 anthropogenic infills. -p re 364 lP 365 5. Discussion 366 In the Iron Age – Phoenician/Persian periods – a dynamic beach environment has been shown on the na 367 western edge of Tel Akko. At the end of the Persian period, the increasing sedimentary inputs led to the ur 368 progressive infilling of the area and Tel Akko lost its access to the sea. Jo 369 5.1. Geomorphological setting of the western edge of Tel Akko 370 Multidisciplinary investigation of the western edge of Tel Akko has demonstrated that this part 371 of the tell has been built on coastal sand dunes covering the kurkar bedrock as shown by the analysis of 372 the ERT profiles and coring. Core AK-XIX-8, drilled along the ERT line Akko-1 (Figure 3), allowed us to 373 demonstrate that the lower resistivity areas in the upper part of the ERT profiles correspond to brown 374 silty-clay sediments with a subsidiary fraction of fine sands, kurkar aggregates and small sherds, most 375 probably related to anthropogenic/agricultural soils. Stratigraphy of the core proved that the high 376 resistivity areas below 6 m depth, reflect the presence of a sand unit. This sand is mainly composed of 377 very fine quartz grains and reflect aeolian deposition of sand in the area. Archaeological evidence 378 confirms the building of city ramparts throughout the tell using this sandy substratum (Figure 4). 379 Analysis of core AK-XIX-5, show the possible presence of a sewage/drainage canal to the sea 380 west of the substantial buildings identified and were where ostraca were found in area G (Figures 9 and 381 12). The presence of the sewage is evidenced by a relatively fine sediment unit containing the ostracod 382 Heterocypris salina which can inhabit brackish temporary water bodies (Figure 9). It has also been 383 demonstrated that H. salina can withstand a certain degree of organic water pollution (Mezquita et al., of 384 1999). Furthermore, we found in this unit numerous pieces of charcoal, grapes seeds and sherds ro 385 indicating the deposition of anthropogenic material originating from the tell during heavy rains in the 386 -p Persian period (Table 2). An even earlier sewage/runoff drainage was found in the southern limit of the re 387 tell, under area PH, dated to the MBII thought at first to be a postern gate (secondary door or gate in a lP 388 fortification; Raban, 1991), but should be identified as a sewage/drainage of the gate area of area P as it 389 was too small to allow the passage of people (Figure 13). na 390 The results of the study of the cores drilled on the western and southwestern areas of Tel Akko outline ur 391 the location the city’s waterfront in the Phoenician/Persian period (Figures 12 and 14). It was evidenced Jo 392 in the form of an exposed shallow sandy to pebbly coastal environment. Rounded sherds of imported 393 ceramics found within the cores, show evidence of weathering by waves in the coastal surf zone and 394 were dated to the same period (Table 3; Supplementary Material). This unit directly covers the substrate 395 composed of kurkar and fine to medium quartz-rich sands probably originating from the weathering of 396 the kurkar sandstone and the deposition of aeolian sands. Considering its elevation in relation to the 397 current sea level and assuming that the relative sea level has been roughly stable since the end of the 398 1st Millennium BC (Sivan et al., 2001), large ships linked to interregional trade could not dock in the 399 immediate proximity of the tell during the Persian period as the water depth would have been <50cm 400 (Figures 10, 12 and 14). Moreover, we may reasonably expect a rapid infill of this area during the Late 401 Persian period because the ceramic sherds discovered are dated from the 4th century BC at the very 402 latest. GPR surveys shows the presence of a aggradational and progradational geomorphic environment 403 such as a coastline in the same area. 404 405 5.2. Location of the anchorage 406 The beach identified at the foot of Tel Akko provided direct access to the open sea but the shallow 407 water depth did not allow for the navigation of large vessels. Around our core AK14-C5 and further of 408 south, the area seems to have been more favourable for the anchorage of large ships during the ro 409 Phoenician and Persian periods (Figure 12). Sediment cores corroborate the interpretation of the ERT 410 -p profiles. While high resistivity units correspond to relatively coarse dry sands and kurkar, lower re 411 resistivity sections (blue colour) are indicative of the presence of fine to medium sands, typical of an lP 412 open marine environment on the southwest of Tel Akko (Figure 5 and 14). Thus, the Phoenician/Persian 413 anchorage, now landlocked, must have existed in the coastal plain located to that southwest area of the na 414 tell (Figure 14). Persian ceramics were previously found in core AK-6 (Morhange et al., 2016; Figure 14) ur 415 at 575 cm depth (200 cm below mean sea level) showing that the water column in the area Jo 416 neighbouring our core AK14-C5 was ca. 200 cm (Figures 11 and 14). This depth was sufficient for the 417 anchorage of maritime vessels (Figure 12). An example of the boats in use along the Levantine coast in 418 the Persian period is given by the Ma‘agan Mikhael wreck, discovered ca. 40 km south of Tel Akko in 419 1985 (Linder and Kahanov, 2003; Kahanov and Linder, 2004). This merchant boat, radiocarbon dated to 420 about 400 years BC, had a draft of 140 cm and a cargo capacity of 15.9 tons (Ben Zeev et al., 2009). We 421 can reasonably assume that this kind of boat was able to approach the coast of Tel Akko with a 200 cm 422 water depth. Moreover, goods could be safely unloaded and transported to the beach by smaller ships 423 such as lighters or barges. Another possibility that cannot be excluded, is the presence of basic 424 constructions of a proto harbour, such as small sized piers, quays or jetties in the area investigated 425 which had facilitated the docking of boats. 426 427 5.3. Coastal progradation and abandonment of Tel Akko 428 Radiocarbon dating and ceramic sherds identification confirms that the sea was present in the 429 immediate proximity of the tell until the 3rd century BC. At that time, the sedimentary inputs on the 430 western side of Tel Akko led to the progressive infilling of the area. The fact that Tel Akko benefited of 431 from a direct access to the sea from the Bronze Age to the Late Persian/Early Hellenistic period has ro 432 largely contributed to its importance in maritime activities over the centuries in the Bay of Haifa/Akko 433 -p and at a larger regional scale (Artzy, 2006). The small apparent sedimentation rate in the area is in re 434 contradiction with the large sedimentation rates evidenced in the centre of the Haifa Bay. Porat et al. lP 435 (2008) have estimated the progradation rate of the bay to be ~40 cm/year between 4000 and 3650 436 years BP (ca. 2000-1650 BC). Afterward, the coastline continued to migrate westwards at an na 437 approximate rate of 50 cm/year until present day. ur 438 In the Akko plain, sediment deposition of that volume reported by Zviely et al. (2007) and Porat Jo 439 et al. (2008) is not apparent. We assume that an obstacle, in the form of an aeolianite ridge (kurkar) 440 identified ca. 400 m to the south of the tell could have blocked the longshore drift and the sedimentary 441 inputs to the north (Figure 14). This ridge probably acted as a natural coastal groyne or dyke. The 442 relative influence of the river and marine sediment flux by longshore drift in the loss of Tel Akko’s direct 443 access to the sea is still difficult to determine precisely. However, our data point toward increasing 444 sedimentary dynamics from marine origin in the southwestern area of Tel Akko leading to the 445 progressive westward progradation of a sandy beach by continuous deposition of coastal sediments 446 during the Late Persian period. 447 Archaeological evidence attest that Tel Akko was occupied until the beginning of the Hellenistic 448 period (Artzy, 2015). Archaeological findings attest to extensive maritime trade reflecting the city’s 449 growth and prosperity until this period (Artzy and Be’eri, 2010). At the same time, the population began 450 relocating to the lower city on the Akko promontory. For a few decades, Tel Akko and the new city were 451 occupied simultaneously (3rd-2nd centuries BC). 452 In our work, we did not find evidence of any coastal or marine environment dated after the Late 453 Persian/Early Hellenistic period. It seems that the main response to the increasing sedimentary input has of 454 been the relocation of the city to the Akko promontory in the Hellenistic period (Figure 2 and 14). Rare ro 455 archaeological finds dated prior to the 3rd century BC and the complete absence of Attic (Greek) 456 -p imported ware, commonly widespread in the coastal region, attest that there was no large-scale re 457 maritime activity at a harbour in the promontory prior to the Hellenistic period (Stern et al., 2008). lP 458 However, from that period, the new city benefited by a commercial (Galili and Rosen, 2018; Galili et al., 459 2010) and a military harbour (Sharvit et al., 2013). Intense maritime activities since the Hellenistic period na 460 were evidences during the dredging of the harbour basin (Galili et al., 2010). These findings show that ur 461 the new city of Akko became the main commercial centre at the expense of the tell. Jo 462 463 6. Conclusion 464 Archaeological excavations have demonstrated that Tel Akko was densely occupied and was a main 465 maritime hub along the coast of the southern Levant during the Phoenician/Persian periods. Intensive 466 maritime activities have been evidenced by the numerous imported ceramics and ostraca discovered on 467 the tell, suggesting that at that time, an anchorage was located in close proximity. Using a 468 multidisciplinary approach to better understand the evolution of the Akko coastal plain to the west and 469 southwest of Tel Akko, we were able to locate the Phoenician/Persian waterfront at the southwestern 470 foot of the tell. We suggest that this area was the most suitable for the anchorage of boats with a water 471 depth reaching ca. 2 m during the Persian period. We recommend future research to be focused around 472 this specific area of the tell. The new data findings highlight the archaeological potential of the coastal 473 plain in immediate proximity to tell and further south-southwest. Local and regional public authorities 474 should be made aware of this new hypothesis in the event of future urban developments planned in the 475 area. 476 477 Author contributions M.G., H.J., Y.S., G.L. and M.A. designed the study. All the authors collected the of 478 data in the field. M.G., L.B. and A.M. carried out laboratory analyses. M.G., H.J., L.B. and M.A. analysed ro 479 the resulting data. M.G., H.J. and M.A. prepared the figures and wrote the original manuscript. All the 480 authors revised the original manuscript draft. -p re 481 lP 482 Data availability Data used for the present research are detailed in the paper and its supplementary 483 files. na 484 ur 485 Acknowledgements Geoarchaeological investigations at Tel Akko have been possible thanks to permits Jo 486 provided by the Israel Antiquities Authority (IAA) and financial support from the BSF (2017-2020). We 487 thank Drs. Nimer Taha and Nicolas D. Waldmann, laboratory manager and laboratory leader of the 488 PetroLab, University of Haifa, for their analytical support and use of their facilities. M.G. thanks the 489 Authority for Advanced Studies and the Leon H. Charney School of Marine Sciences at the University of 490 Haifa and the BSF for the funding of his post-doctoral fellowship. H.J. thanks the University of Wisconsin 491 - Eau Claire’s Office of Research and Sponsored Programs for Student/Faculty Research, Collaboration 492 and Diversity mentoring Grants for Logan Bergevin and Ethan Sailer-Haugland. MA is thankful to the 493 Hatter Foundation and an anonymous donor. We thank two anonymous reviewers for their careful 494 reading of our manuscript and their insightful comments and suggestions. 495 496 Funding This research has been carried out thanks to the support of the United States-Israel Binational 497 Science Foundation (BSF) to the project “A Palaeo-Hydrological Approach to Understanding the Complex 498 Coastal Evolution of an Ancient Maritime Trade City: The Tel Akko Multi-Disciplinary Case Study, Israel” 499 to M. Artzy and H.M. Jol (Grant No. 2016080). 500 501 References of 502 Artzy, M. (2006) The Carmel Coast during the second part of the Late Bronze Age: A center for Eastern ro 503 Mediterranean transhipping. Bulletin of the American Schools of Oriental Research 343(1), 45-64. 504 -p Artzy, M. (2012) Return to Tel Akko, its anchorages, harbour, and surroundings. Recanati Institute for re 505 Maritime Studies Newsletter, University of Haifa, 37, 5-14. lP 506 Artzy, M. (2015) What is in a name? ‘Akko–Ptolemais–‘Akka–Acre. Complutum 26(1), 205-212. 507 Artzy, M., Be’eri, R. (2010) Tel Akko. In A.E. Killebrew and V. Raz-Romero (eds.), One thousand nights na 508 and days—Akko through the ages, 14-23. Hecht Museum: Haifa. ur 509 Artzy, M., Quartermaine, J. (2014) How and When Did Tel Akko Get Its Unusual Banana Shape? In Y. Jo 510 Galanakis, T. Wilkinson and J. Bennet (eds.) Αθυρματα. Critical essays on the archaeology of the eastern 511 mediterranean in honour of E. Susan Sherrat, 11-22. Archaeopress Archaeology: UK. 512 Artzy, M., Jol, H., Giaime, M., Salmon, Y., abu-Hamid, A., López, G.I., Morhange, C., Kaniewski, D., 513 Bauman, P., Killebrew, A.K. (2021) Coastline, River Changes and their Effects on Anchorages / Harbors 514 and Habitation Patterns: Akko (Israel) as an Example. In S. Demisticha and L. Blue (eds), Under the 515 Mediterranean I. Studies in Maritime Archaeology, 267-278. Sidestone Press, Leiden, Netherlands. 516 Athersuch, J., Horne, D.J., Whittaker, J.E. (1989) Marine and brackish water ostracods (Superfamilies 517 Cypridacea et Cytheracea): Keys et notes for the identification of the species. E.J. Brill: Leiden, The 518 Netherlands:. 519 Avnaim-Katav, S., Agnon, A., Sivan, D., Almogi-Labin, A. (2016) Calcareous assemblages of the 520 southeastern Mediterranean low-tide estuaries. Seasonal dynamics and paleo-environmental 521 implications. Journal of Sea Research 108, 3049. 522 Beeri, R. (2008) Tel Akko and the Urbanization of Akko plain in the first half of the second Millennium 523 BCE. Unpublished PhD thesis, Department of Archaeology, University of Haifa. (In Hebrew) 524 Ben Zeev, A., Kahanov, Y., Tresman, J., Artzy, M. (2009) The Ma‘agan Mikhael Ship: A Reconstruction of 525 the Hull. Israel Exploration Society and Leon Recanati Institute for Maritime Studies, University of Haifa: of 526 Haifa. ro 527 Conyers, L.B., Goodman, D. (1997) Ground-penetrating radar: An Introduction for archaeologists. 528 AltaMira Press. -p re 529 D’Angelo, G., Garguillo, S. (1978) Guida alle Conchiglie Mediterranee, conoscerle, cercarle, collezionarle. lP 530 Fabri Editori: Milan. 531 Dothan, M. (1976) Akko: Interim excavation report first season 1973/74. Bulletin of the American na 532 Schools of Oriental Research 224, 1-48. ur 533 Dothan, M. (1985a) A Phoenician Inscription from Akko. Israel Exploration Journal 35(2/3), 81-94. Jo 534 Dothan, M. (1985b) Ten Seasons of Excavations in Ancient Akko. Qadmoniot 18, 2-24. 535 Elyashiv, H., Bookman, R., Zviely, D., Avnaim-Katav, S., Sandler, A., Sivan, D. (2016) The interplay 536 between relative sea-level rise and sediment supply at the distal part of the Nile littoral cell. The 537 Holocene 26(2), 248-264. 538 Empereur, J.Y. (1995) Le Port Hellénistique d’Amathonte. In V. Karageorghis and D. Michaelides, D. 539 (eds.) Cyprus and the Sea Proceedings of the International Symposium, 131-138. University of Cyprus- 540 Cyprus Ports Authority: Nicosia. 541 Empereur, J.Y., Verlinder, C. (1987) The Underwater Excavations at the Ancient Port of Amathus in 542 Cyprus. International Journal of Nautical Archaeology 16, 7-18. 543 Empereur, J.Y., Koželj, T., Picard, O., Wurch-Koželj, M. (2018) The Hellenistic Harbour of Amathus: 544 Underwater Excavations, 1984-1986. Volume 1, Architecture and History. Études Chypriotes (19), École 545 Française d’Athènes: Athens. 546 Frenzel, P., Boomer, I. (2005) The use of ostracods from marginal marine, brackish waters as 547 bioindicators of modern and quaternary environmental change. Palaeogeography, Palaeoclimatology, 548 Palaeoecology 225(1), 68-92. 549 Galili, E., Rosen, B. (2017) The Akko Marina Archaeological Project—Summary. In E. Galili (Ed.), The Akko of 550 Marina Archaeological Project, 320-344. BAR International Series 2862. BAR Publishing: Oxford, England. ro 551 Galili, E., Rosen, B., Zviely, D., Silberstein, N., Finkielsztejn, G. (2010) The evolution of Akko harbour and 552 -p its Mediterranean maritime trade links. Journal of Island and Coastal Archaeology 5, 191-211. re 553 Gambash, G. (2014) En route to Egypt: Akko in the Persian period. Journal of Near Eastern Studies 73(2), lP 554 273-282. 555 Giaime, M., Morhange, C., Marriner, N., López, G.I., Artzy, M. (2018) Geoarchaeological investigations at na 556 Akko, Israel: New insights into landscape changes and related anchorage locations since the Bronze Age. ur 557 Geoarchaeology 33(6), 641-660. Jo 558 Giaime, M., Marriner, N., Morhange, C. (2019) Evolution of ancient harbours in deltaic contexts: A 559 geoarchaeological typology. Earth-Science Reviews 191, 141-167. 560 Inman, D. L. (2003) Littoral cells. In M. Schwartz (Ed.), Encyclopedia of coastal science. The Earth Sciences 561 Encyclopedia Online, 594-599. Kluwer Academic Publishers: Dordrecht, Netherlands. 562 Jol, H.M., Bristow, C.S. (2003) GPR in sediments: Advice on data collection, basic processing and 563 interpretation, a good practice guide. In C.S. Bristow, H.M. Jol (eds), Ground penetrating radar in 564 sediments, 9-27. London: Geological Society, Special Publications, 211. 565 Kafri, U., Ecker, A. (1964) Neogene and Quaternary subsurface geology and hydrogeology of the Zevulun 566 Plain. Ministry of Development, Geological Survey: Israel. 567 Kahanov, Y., Linder, E. (2004) The Ma‘agan Ship: The Recovery of a 2400-Year-Old Merchant-man, 568 Volume II. J. Tresman (ed), Israel Exploration Society and University of Haifa: Jerusalem. 569 Kaniewski, D., Van Campo, E., Morhange, C., Guiot, J., Zviely, D., Shaked, I., Otto, T., Artzy, M. (2013) 570 Early urban impact on Mediterranean coastal environments. Scientific Reports 3, 3540. 571 Lachenal, A.M. (1989) Écologie des ostracodes du domaine méditerranéen: Application au golfe de 572 Gabès (Tunisie orientale), les variations du niveau marin depuis 30 000 ans. Documents des Laboratoires 573 de Géologie de Lyon, 108. of 574 Linder, E. and Raban, A. (1965) Underwater survey of Akko Harbour. In The Western Galilee and the ro 575 Galilee Coast. The 19th Convention for Yediat Haaretz, 180–193. Jerusalem. (In Hebrew) 576 -p Linder, E., Kahanov, Y. (2003) The Ma‘agan Mikhael Ship: The Recovery of a 2400-Year-Old Merchant- re 577 man, Volume I. E. Black (ed), Israel Exploration Society and University of Haifa: Jerusalem. lP 578 Marriner, N., Morhange, C., Kaniewski, D., Carayon, N. (2014) Ancient harbour infrastructure in the 579 Levant: tracking the birth and rise of new forms of anthropogenic pressure. Scientific Reports 4, 5554. na 580 Marriner, N., Morhange, C., Flaux, C., Carayon, N. (2017) Harbors and Ports, Ancient. In A.S. Gilbert, P. ur 581 Goldberg, V.T. Holliday, R.D. Mandel, R.S. Sternberg (eds.) Encyclopedia of Geoarchaeology, 382-403. Jo 582 Springer Netherlands: Dordrecht. 583 Mezquita, F., Griffiths, H., Sanz, S., Soria, J., Piñón, A. (1999) Ecology and Distribution of Ostracods 584 Associated with Flowing Waters in the Eastern Iberian Peninsula. Journal of Crustacean Biology 19(2), 585 344-354. 586 Morhange, C., Giaime, M., Marriner, N., abu Hamid, A., Bruneton, H., Honnorat, A., Kaniewski, D., 587 Magnin, F., Porotov, A.V., Wante, J., Zviely, D., Artzy, M. (2016) Geoarchaeological evolution of Tel 588 Akko’s ancient harbour (Israel). Journal of Archaeological Science: Reports 7, 71-81. 589 Pérès, J.M. (1982) Major Benthic Assemblages. In O. Kinne (ed.), Marine Ecology, Part 1 vol. 5, 373-522. 590 Wiley: Chichester. 591 Pérès, J.M., Picard, J. (1964) Nouveau Manuel de Bionomie Benthique de La Mer Méditerranée. Periplus: 592 Marseille. 593 Porat, N., Sivan, D., Zviely, D. (2008) Late Holocene embayment infill and shoreline migration, Haifa Bay, 594 Eastern Mediterranean. Israel Journal of Earth Sciences 57(1), 21-31. 595 Raban, A. (1983) The Biblical Port of Akko on Israel’s Coast. Archaeology 36(1), 60-61. 596 Raban, A. (1985) The ancient harbours of Israel in Biblical times. Harbour Archaeology, Proceedings of 597 the First International Workshop, 11-44. BAR Publishing: Oxford, England. of 598 Raban, A. (1991) The port city of Akko in the MBII. Michmanim 5, 17-34. ro 599 Raban, A. (1993) A group of imported ‘East Greek’ Pottery from Locus 46 at Area F on Tel Akko. In M. 600 -p Heltzer, A. Segal, D. Kaufman (eds.) Studies in archaeology and History of Ancient Israel, in Honour of re 601 Moshe Dothan, 73-98. University of Haifa: Haifa. lP 602 Raban, A., Stieglitz, R.R. (1993) Phoenicians on the Northern Coast of Israel in the Biblical Period. Reuben 603 and Edith Hecht Museum, University of Haifa: Haifa. na 604 Reimer, P. J., Austin, W. E. N., Bard, E., Bayliss, A., Blackwell, P. G., Bronk Ramsey, C., Butzin, M., Cheng, ur 605 H., Edwards, R. L., Friedrich, M., Grootes, P. M., Guilderson, T. P., Hajdas, I., Heaton, T. J., Hogg, A. G., Jo 606 Hughen, K. A., Kromer, B., Manning, S. W., Muscheler, R., Palmer, J. G., Pearson, C., van der Plicht, J., 607 Reimer, R. W., Richards, D. A., Scott, E. M., Southon, J. R., Turney, C. S. M., Wacker, L., Adolphi, F., 608 Büntgen, U., Capano, M., Fahrni, S. M., Fogtmann-Schulz, A., Friedrich, R., Köhler, P., Kudsk, S., Miyake, 609 F., Olsen, J., Reinig, F., Sakamoto, M., Sookdeo, A. Talamo, S. (2020) The IntCal20 Northern Hemisphere 610 radiocarbon age calibration curve (0-55 cal kBP). Radiocarbon, 1-33. 10.1017/RDC.2020.41 611 Sharvit, J., Planer, D., Buxton, B. (2013) Preliminary findings from archaeological excavations along the 612 foot of the southern seawall of Akko, 2008–2012. Michmanim 24, 39-40. (In Hebrew with English 613 abstract) 614 Sivan, D., Gvirtzman, G., Sass, E. (1999) Quaternary stratigraphy and paleogeography of the Galilee 615 coastal plain, Israel. Quaternary Research 51(3), 280-294. 616 Sivan, D., Wdowinski, S., Lambeck, K., Galili, E., Raban, A. (2001) Holocene sea-level changes along the 617 Mediterranean coast of Israel, based on archaeological observations and numerical model. 618 Palaeogeography, Palaeoclimatology, Palaeoecology 167(1-2), 101-117. 619 Stanley, J. D., Bernhardt, C. E. (2010) Alexandria’s Eastern Harbor, Egypt: Pollen, microscopic charcoal, 620 and the transition from natural to human-modified basin. Journal of Coastal Research 26(1), 67-79.. of 621 Stern, E., Galili, E., Rosen, B. (2008) Acco. In E. Stern (ed.), The New Encyclopedia of Archaeological ro 622 Excavations in the Holy Land, 24-36. Israel Exploration Society: Jerusalem; Biblical Archaeology Society: 623 Washington, DC. -p re 624 Stuiver, M., Reimer, P.J. (1993) Extended 14C database and revised CALIB radiocarbon calibration lP 625 program. Radiocarbon 35, 215-230. 626 Stuiver, M., Reimer, P.J., Reimer, R.W. (2020) CALIB 8.2 [WWW program] at http://calib.org, accessed na 627 2020-8-17. ur 628 Thompson, R. Oldfield, F. (1986) Environmental Magnetism. George Allen & Uniwin: London. Jo 629 Vachtman, D., Sandler, A., Greenbaum, N., Herut, B. (2013) Dynamics of suspended sediment delivery to 630 the Eastern Mediterranean continental shelf. Hydrological Processes 27, 1105-1116. 631 Zviely, D, Sivan, D, Ecker, A., Bakler, N., Rohrlich, V., Galili, E., Boaretto, E., Klein, M., Kit, E. (2006) 632 Holocene evolution of the Haifa bay area, Israel, and its influence on ancient tell settlements. The 633 Holocene 16, 849-861. 634 Zviely, D., Kit, E., Klein, M. (2007) Longshore sand transport estimates along the Mediterranean coast of 635 Israel in the Holocene. Marine Geology 237, 61-73. 636 637 Figure captions/Table captions 638 639 Figure 1: A) General location of Haifa Bay in the southern Levant (after Inman, 2003; image ©CNES). 640 Dominant wind and wave patterns at Alexandria are shown in the rose diagrams (Stanley and Bernhardt, 641 2010). B) Geomorphological reconstruction of Haifa Bay 4000 years BP (after Elyashiv et al., 2016). 642 643 Figure 2: Location of the main archaeological areas and main habitation centres cited in the text. 644 Background image © Zoom Earth, Microsoft. 645 of 646 Figure 3: A) Location map of ERT lines collected on Tel Akko southern plain (background image taken by ro 647 drone © P. Bauman, 2018). B) Location map of the sedimentary cores, GPR grids (to add), ERT profiles 648 -p and excavated areas at Tel Akko. Start of the ERT line (0m) is shown by the yellow dot (background re 649 image taken by drone © J. Quatermaine, 2019). lP 650 651 Figure 4: ERT lines Akko-1, Akko-2 and Akko-3. The profiles were collected on the western edge of Tel na 652 Akko (see Figure 3). Horizontal scale is 9.42 pixels per unit spacing. Vertical exaggeration = 1.23. High ur 653 resistivity features attest of the presence of sands/kurkar sandstone. Jo 654 655 Figure 5: ERT lines Akko-14, Akko-16 and Akko-17 collected on the Tel Akko coastal plain (see Figure 3). 656 Most of the sections are dominated by low resistivity (blue) material which are interpreted as fine over 657 saturated sediments (silty-clay and fine sands) 658 659 Figure 6: Ground penetrating radar data was collected on the western portion of the tell (near Td 660 excavation). The upper plan view image is a depth slice from 0.44 m to 0.55 m depth. A curvilinear 661 feature is highlighted in this depth horizon. The lower image is Line 9 of the data collected. From 0 m to 662 12 m, multiple subhorizontal to horizontal continuous reflections become dipping reflections at 12 m to 663 20 m. 664 The reflection pattens from the GPR data collected are interpreted as an aggradational and 665 progradational geomorphic environment such as a coastline. 666 667 Figure 7: Core AK-XIX-8. This sediment core shows the presence of two distinct stratigraphic units. See 668 Figure 3 for location of the core. of 669 ro 670 Figure 8: Cores AK-XIX-6 and AK-XIX-7. The sediment cores show the presence of an supralittoral 671 -p environment followed by the development of an agricultural soil. See Figure 3 for location of the cores. re 672 lP 673 Figure 9: Sedimentary texture, magnetic susceptibility and analysis of ostracods in core AK-XIX-5. Three 674 different environmental units were identified. The unit highlighted in light brown show the na 675 sewage/drainage identified in this core. Fauna is dominated by the ostracod specie Heterocypris salina. ur 676 See Figure 3 for location of the core. Jo 677 678 Figure 10: Sedimentary texture, magnetic susceptibility and analysis of ostracods of cores AK-XIX-3, AK- 679 XIX-4, AK-XIX-9 and AK-XIX-10. The different environmental units identified shows the presence of a 680 shallow coastal environment dated to the Persian period (units highlighted in blue). See Figure 3 for 681 location of the cores. 682 683 Figure 11: Sedimentary texture, magnetic susceptibility and analysis of ostracods of cores AK14-C5 and 684 AK17-C7. The different environmental units identified are characteristic of a coastal progradation. The 685 infralittoral environment in which we identified marine and coastal ostracods is followed by the 686 development of a supralittoral beach environment barren of fauna. See Figure 3 for location of the 687 cores. 688 689 Figure 12: Correlation of sediment cores retrieved from the west and southwest feet of Tel Akko. 690 Location of transects is given in Figure 3. These transects show the geomorphology of the beach 691 environment and the variation of the water column height. The different shades of blue depict the 692 infralittoral zone (always submerged) from the mediolittoral zone (which corresponds to the transition of 693 zone between a permanently emerged and a permanently submerged environment). C14 dating of ro 694 cores AK-XV-1 and AK-XV-2 (Giaime et al., 2018) have been calibrated using the new calibration curve 695 -p IntCal20 (Reimer et al., 2020). See Figure 3 for the location of the cores. re 696 lP 697 Figure 13: Photo of the MBII (2000-1750 BC) postern gate in area P, from the south. Photo © M. Artzy 698 na 699 Figure 14: Palaeoenvironmental reconstruction of Tel Akko maritime interface in the Phoenician/Persian ur 700 period. The different environments have been identified based on this study and from previous Jo 701 research (Morhange et al., 2016; Giaime et al., 2018). The dotted line shows the location of the core 702 transects as seen in Figure 12. Topography was adapted from a map produced by Treidel (1925-1926). 703 704 Table 1: Main historical periods of Tel Akko from the Bronze Age to the Hellenistic period. 705 706 Table 2. AMS-14C data expressed in calibrated years BP and BC at the 95% confidence level (2σ). using 707 Calib 8.2 (Stuiver and Reimer, 1993; Stuiver et al, 2020) and the IntCal20 curve (Reimer et al., 2020). 708 amsl: above mean sea level. 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 of ro -p re lP na ur Jo Sample Lab Material Depth (cm Elevation Age 2 sigma 2 sigma cal. Historical number below (cm amsl) 14C cal. BP BC/AD min; Period surface) min; max max AK-XIX-3 ULA- Charcoal 578 45 2800 2850 ; 1010 ; 900 Early Iron 578 8700 amsl ± 20 2960 BC Age AK-XIX-4 ULA- Charcoal 565 106 amsl 2455 2370 ; 750 ; 420 Late IA - 565 8695 ± 15 2700 BC Persian AK-XIX-5 ULA- Charcoal 678 141 2455 2370 ; 750 ; 420 Late IA - 678 8698 amsl ± 20 2700 BC Persian AK-XIX-5 ULA- Seed 710 106 amsl 2485 2470 ; 770 ; 520 Late IA - 710 8697 (grape) ± 20 2720 BC Persian of ro -p re lP na ur Jo Declaration of interests ☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. ☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: of ro -p re Sincerely, On behalf of all authors lP Dr. Matthieu Giaime na Haifa, August 2020. ur Jo