Jezreel Valley Regional Project Archaeology and History of a Regional Landscape

2011 Summer Season at Tel Megiddo East

2011 Summer Season at Tel Megiddo East

Matthew J. Adams (Bucknell University)

Jonathan David (Gettysburg College)

Robert Homsher (University College London)

Margaret Cohen (The Pennsylvania State University)


with additional data and input by

Melissa Cradic (University of California, Berkeley)


Version: 10 April 2013

The 2011 Excavation Season

This season was the third excavation season following the 2010 exploratory excavations at ‘Ain el-Qubbi and Tel Megiddo East and the Ground Penetrating Radar survey at Tel Megiddo East and Legio in December of 2010. The 2011 season concentrated excavation on the site of Tel Megiddo East, with additional geoarchaeological and geophysical studies in adjacent areas and at the nearby site of the Roman legionary camp at Legio. The 2011 season afforded opportunities to pursue a broad research agenda on the Early Bronze Age, to assess the viability of a number of geophysical techniques for future work, to conduct a geoarchaeological study of the paleoenvironment around Megiddo, and to establish methodologies for the initiation of a valley-wide survey project. Our goals for the season were the following:

 1. The JVRP aims to document the broader EB I settlement associated with Tel Megiddo. From 1996-2010, The Tel Aviv University Megiddo Expedition uncovered an unprecedentedly large temple and cultic compound dating to the EB I (Adams, Finkelstein, and Ussishkin forthcoming). This Great Temple is now well-known, but the larger settlement landscape, the home of its builders, has never been explored. On the basis of previous surface surveys (Finkelstein et al. 2006, see above) and geophysical surveys (Eppelbaum and Itkis 2000; and the 2010 GPR survey, above), the JVRP selected two areas associated with a small mound directly east of and downslope from Tel Megiddo for excavation, Area A and Area B (Figs. 1-2). The location of this mound and the positive results from the geophysical surveys suggested that this small mound had potential for the location of the EB I town (see above). Additionally, Area C was selected as a major area of investigation on the basis of eye-witness reports, verified by our team, of exposed architecture on the surface.

2. In consideration of future research and surveys of other sites in the Jezreel Valley, the project team is interested in testing a variety of geophysical remote sensing technologies. The 2010 GPR survey pointed to several anomalies within the mound on the east side of the tel (see above). Our 2011 excavation Areas A and B were situated to investigate a number of these anomalies in order to check the reliability of GPR within the geological character of this region. Additional GPR work conducted in 2010 by the JVRP at Legio, in collaboration with Yotam Tepper of the IAA, afforded another opportunity to test the usefulness of GPR. In 2011, the JVRP also conducted an Electromagnetic Survey (EM) at the same location as the Legio GPR survey in order to compare the results of these two technologies. In future seasons, the project team aims to assess the reliability and cost-effectiveness of both technologies by excavating in this area.

 3. Within the framework of the JVRP’s landscape archaeology approach to studying areas of human activity, both within and between settlements, the team is also interested in identifying localities of activity that are associated with but not contiguous with settlements themselves (e.g., the quarry identified in 2010).

 4. Recognizing that the environment is an essential part of how humans conduct their daily business and how they form personal and group identities, the JVRP is now developing methods for the acquisition of geological and paleoenvironmental data. The Jezreel Valley Regional Project, in cooperation with the Tel Aviv University Megiddo Expedition, is engaged in a multi-faceted research program which aims to contextualize the archaeological discoveries from Tel Megiddo with paleo-environmental data, settlement evidence, and landscape research from areas surrounding the tel. While there is no limit to our chronological interest, our initial focus is on the Early Bronze Age I and the Roman Period, two eras for which there is the most abundant information in this vicinity. In 2011, the team selected seven locations for coring with a mechanical backhoe (Fig. 3). The following research questions were instrumental in determining the purpose and location of these backhoe (JCB) cores: (A) reconstructing the pre-medieval topography by documenting the (post-)Byzantine colluvial accumulation; (B) testing for settlement activity outside the boundaries of the tel; (C) acquiring long term paleoenvironmental data via pollen extracted from stream-bed sediment; (D) assessing the long term flow and deposition of ‘Ain el-Qubbi with both geophysical (GPR) and stratigraphic analysis of the stream-bed; (E) acquiring additional geomorphological data in order to reconstruct topography and access to natural resources over time.















Excavation Areas A and B (2011)

Areas A and B yielded similar results over the course of two weeks in the summer 2011 season, providing paleoenvironmental data and negative evidence regarding the location of the EB I settlement. Area B was situated on a gentle rise of an agricultural (watermelon) field on the east side of the highway (Road 66) at the eastern base of Tel Megiddo (Figs. 1-2). The location of the area was determined by its proximity to the tel and its position over a cluster of anomalies presented in the data from 2010 GPR survey. These particular anomalies appeared in the survey Grid 2 (Figs. 2) at varying depths of 0.6 – 1.8 m beneath the surface. Therefore, the squares in Area B were placed in order to maximize coverage of these anomalies.

 Beneath the Area B topsoil was a mixed matrix of coarse pebbles and pottery sherds. The pebble matrix was found to consist of multiple layers of deposition with increasingly silty sediment towards the lowest extent of the pebbles and ultimately underlying the pebbles. These deposits appeared in a number of units (B-5, C-5, D-5, D-6, C-9, C-10, and D-10; see Figs. 2-5), but in most cases, not encountered in the entire square. In a number of units a pattern emerged indicating that these pebble/sherd loci were channels aligned in a general northwest to southeast direction. The team sectioned one of these channels to study the depositional activity within it (Fig. 5).

In every instance where the pebble and silt layers were removed, excavation progressed downward into a layer of very dense, dark greyish-brown clay. Aside from a few small sherds that presumably resulted from later disturbance (e.g., roots, animal holes, plow), this clay material was sterile and continued deeper than the extent of excavation in Area B. Wherever the pebble and silt material was not encountered in the area, the layer of topsoil gradually gave way to this underlying clay. Segregating the topsoil/plow zone from the underlying clay was difficult since both sediments are essentially the same in composition, suggesting that the clay material extended to the surface where it has since been utilized for agriculture. Likewise, many small stones characteristic of the pebble layer were scattered about the surface of the field.

Excavation continued in order to remove completely the material related to the pebble and silt deposit, and to reach GPR anomaly depths in the underlying clay. Excavation ceased in the area once the GPR depths had been reached in a number of units. No architecture or occupational materials were encountered in any squares, and it was determined that the underlying clay material found in each square represented geological deposition pre-dating human settlement in the area. These channels appear to be the result of seasonal drainage from the vicinity of Tel Megiddo southward and eastward into the local wadi system that is still visible today (Fig. 6). The implications of these findings will be discussed below.

Area A was located 50 m north of Area B in a low-lying area just north of the low rise upon which Area B sits (Figs. 1-2). Area A was sited where the confluence of the GPR data (see above) and the magnetometer data (Eppelbaum and Itkis 2000) suggested that human activity could be detected. The four western squares of the area encountered silty clay deposits immediately beneath the plow zone (Fig. 7). The silt contained some sparse cultural material including Roman and Byzantine sherds, usually small and worn, a few coins, and an occasional EB I sherd. In most units, excavation ceased at approximately one meter in depth, but unit E-7 continued to a depth of two meters down to the marl of the Tayiqe formation (also found in the ‘Ain el-Qubbi excavations to the northeast). Near the bottom of this excavation, some 20 cm above the marl, the team encountered a thin silty layer containing a few small EB I sherds and some bone material. The deposition of this EB I material appears to be similar to the deposition of the rest of the silt, formed in the context of a marshy environment. Most of the two meters of silt and clay excavated can be associated with marsh-like conditions in the vicinity of the spring, including drainage from at least the Early Bronze Age through the Roman periods (Figs. 6-8).

In two units on the west side of the area, excavation revealed a broad, shallow pit containing several discarded blocks, several late Roman coins, and largely Roman period pottery, with a few small Byzantine and Islamic sherds. The pit was cut into the calcareous Pleistocene clay deposit, the same layer encountered in Area B.

Both Areas A and B demonstrate that the space immediately east of Tel Megiddo and of the modern road was uninhabited in antiquity. Area A was swampy at least through the Roman period, and Area B (at least in the Chalcolithic and Early Bronze I) was prone to seasonal erosion For further interpretation of the geomorphological evidence, see below.



JCB Coring

The seven trenches (counting right to left, according to Fig. 3) excavated in the JCB coring project provided a number of insights into the paleoenvironment and the potential extent of the settlement:

JCB Trench 1: This core was located to acquire pollen samples for a long-term palynological study by Dafna Langgut of Tel Aviv University. These samples are still being processed. The upper 70 cm had small Roman/Byzantine sherds in lacustrine deposits; below that was pure Pleistocene clay.

JCB Trench 2: This unit tested what appeared to be a sprawl of pottery, perhaps indicating an archaeological site. The trench showed that the first 80 cm below the surface contained EB sherds in a colluvial matrix, beneath which was Pleistocene clay. The colluvial material was probably eroded from the general area of 2011’s excavation Area C.

JCB Trench 3: This core was designed to be a preliminary examination of the wadi channel into which the spring empties. This trench found that the ancient channel has seen significant filling since the Hellenistic period. We found four meters of colluvial and alluvial deposits filled the channel,above pre-Hellenistic material. A Hellenistic coin was found at the maximum depth of the trench at 4 m below the surface (see discussion below).

JCB Trench 4: This test was important because an IAA JCB trench only 50 m to the northeast indicated stratified cultural deposits from the EB through the Persian period. However, our excavation here showed some EB I sherds in the first meter, apparently deposited in a colluvial context. Beneath this was sterile clay capping the calcareous marl found in the ‘Ain el-Qubbi test excavations.

JCB Trenches 5 and 6 both revealed sterile Pleistocene clay immediately beneath the plow zone.

JCB Trench 7 found bedrock beneath Pleistocene clay at a depth of 1 m.


Excavation Area C (2011)

With the excavation of Area C, the team located a significant portion of the Early Bronze I settlement. Area C was located along the western slope of a hill ca. 325 m to the east of the highway at the eastern base of Tel Megiddo. This hill is presumably representative of an outcrop of bedrock on the opposite side of the lower-lying topography from the tel, based on its relatively higher elevation and portions of bedrock being visible on its surface at some points. Furthermore, the bedrock within this vicinity is known to be at a shallow depth from the present surface based on the 2010 investigations of Tel Megiddo East in the adjacent agricultural field. During the 2010 season, evidence of EB Ib occupation was found in unit BB-104 (located ca. 60 m north of Area C) nestled against the edge of the bedrock that dropped in elevation at that particular point. The location of Area C was based on three primary factors: (1) the assumption that EB I occupation in the valley might be directly related to the exploitation of natural bedrock, as known from the Megiddo “Stages” and unit BB-104; (2) as of yet, there is no evidence for any occupation in the lower elevations in the valley; (3) the top of a wall had previously been observed in the southern drainage ditch of the dirt road running along the southern end of the aforementioned agricultural field, prompting the establishment of excavation units upon a known architectural element.


Area C Phasing

 Due to the discrepancy in elevation (ca. 3 m) along the slope of Area C, it was difficult to tie together phases of architecture throughout the entire area, especially since there were no floors associated with almost any of the walls. Nonetheless, there appear to be four main architectural phases in the area. The following descriptions of phases form a preliminary stratigraphy of Area C.

Phase I, the earliest phase, is demonstrated by W22 in D-5 and W23 in E-4. W22 is a curvilinear wall of one course and unknown width extending partly into the square from the north section. This wall’s top elevation (111.23) is the lowest of all architectural elements in the area, and its northern extension was not found in the probe (L.070) in D-4. W23 appears to be a straight wall 0.45 m wide, orientated north-south and appearing to run beneath W13; however, the northern extension of this wall was not found in probe L.077. These two walls, which were the lowest encountered, lie beneath other architectural elements and do not align with the orientation of later phases.

Phase II is demonstrated by W13 in E-4 and W07 in D-4. Both of these walls were ca. 0.5 m wide and constructed of cobbles, which is quite a distinct type of construction from all other architecture in the area. W13 was preserved in two courses and was curvilinear, forming a sort of horseshoe, with what seemed to be an entrance or opening in the southeast. Associated with this wall was a partial surface (L.069) inside the wall, including a flat basalt slab that seemed to be situated somewhat centrally inside the presumed boundary of the wall. W07 was straight, extending east-west through the square, with only one course preserved (and some disturbance from later walls). Associated with this wall were surfaces L.014, including basalt Pavement 025 and Vessel 1, to the south and L.050 to the north. Above L.050, L.002 had articulated two flat stone slabs, which may have been pillar bases; the lowest (and south) of these slabs (and most rounded) possibly relates to W07 and architectural phase II.

Phase III consists of two sub-phases, IIIA and IIIB, which essentially represent original wall constructions and subsequent rebuilds. Phase IIIA, the earlier of these two sub-phases, is demonstrated by: W29 in E-5; W26 and W30 in F-4; W25 in F-5; W20, W27 and W28 in G-4; W12 in G-5; W19 in H-4; and W21 in H-5. W29 appears to be a southern continuation of W26 and W30 in F-4, and is the lowest of a series of rebuilds. It consists of two main rows and is ca. 0.8 m wide with small stones abutting to the east, and possibly intentional rubble fill to the west (L.072). W30, to which W26 appears to belong, is of a similar width to W29 and runs in a northern continuation of that same orientation. W25 consists of two to three courses of large stones on its western face and rubble on its eastern side, the latter of which abuts the sloping bedrock as a sort of revetment. Although the northern boundary of this wall terminates as it hits the bedrock in the north of the square, a possible reconstruction of the southern portion of W28 in G-4 could be a northern continuation of the orientation. W28 appears to either make a corner with W20 or continue northward with W27; in any case, the southern extent has probably been disturbed. W20 is the western continuation of east-west W19 in H-4, both of which appear only to have a well-constructed northern face with more of a rubble construction to the south. This construction may be similar to that of W25, assuming there is underlying bedrock to the south. More than any other wall, W19 retained at least one course of light grey mud-brick in the southeast. W12 in G-5 and W21 in H-5 belong to the same east-west orientated wall, which is ca. 1.05 m wide in three to four rows of medium-sized stones and two courses high.

Phase IIIB is demonstrated by walls: W16 and W24 in E-5; W15, W17 and W18 in F-4; W14 in F-5; and W10 in G-4. W16 measured ca. 0.6 m wide and was built above, but slightly offset to the west of, W29, which extended further east. W24 was of slight construction, consisting of two rows and one course of small stones extending perpendicularly westward from W16. W15, which appeared to be the northern continuation of W16, was ca. 0.8 m wide and ran diagonally, bisecting the square. Abutting W15 perpendicularly from the west were walls W17 (south), of one row of medium-sized stones, and W18 (north), ca. 0.8 m wide with two rows of large stones. W14 is situated directly on W25, which seems to serve as a terrace, and measures ca. 0.95 m wide with three rows and one course. W10 was constructed (probably as a revetment of some sort) above the southern side of W20 and consisted of at least five courses of cobble-sized stones leaning slightly northward. This curious construction may have been supporting some sort of rubble or earthen fill subsequently above W20.

In square D-4, it remains difficult to accurately situate walls W01, W08 and W09 in the overall phasing of the area. However, Phase III seems to be most probable fit since it follows directly in sequence from Phase II, just as W01, W08 and W09 were built directly on W07. W01 measured 0.7 m wide with two rows and curved from the west to the south, entering and exiting the square. W08 was fragmentary and consisted of only one row of medium-sized stones orientated east-west. W09 was also fragmentary and consisting of only one row of stones but oriented north-south next to W01. W08 and W09 both disturbed the portions of W07, on which they were situated. The higher (and north) of the two flat stone slabs may relate to this phase.

Phase IV, which generally consisted of only segments of walls, is demonstrated by walls: W03 in D-5; W02 in E-4; W11 in E-5; W04 and W05 in F-4; and W06 in G-4. W03 measured 0.45 m wide and was orientated east-west, running through the northeast corner of D-5. W02 measured ca. 0.8 m wide and ran east-west through the northeast of E-4, but was very disturbed. W11 consisted of only a few medium- to large-sized stones above W16, and offset to the east from W16. W04 measured 0.65 m wide and ran east-west in the northeast corner of F-4 above W18 and W15, offset to the north of W18 by one row. W05 also measured 0.65 m wide and consisted of a segment orientated north-south built above W15, but offset by one row to the east. W06 measured 0.6 m wide and ran north-south through the northwest corner of G-4, possibly making a corner with W04 in baulk F/G-4.


Area C Summary and Observations

Based on the architectural phases and patterns based on ceramic observations, there appears to be a general trend in the EB I settlement activity in Area C. The earliest architectural Phases I and II only appear furthest downslope and in squares with no evidence of bedrock. The architecture demonstrated during these phases is relatively heterogeneous, with thinner walls than later. Beginning in Phase III, construction expands further upslope onto the bedrock with much larger-scale walls than before and terracing revetments built against the bedrock. The widths of walls also become more homogeneous, as well as their linear orientation across the entire area. At this same time, occupation seems to taper off in the lowest squares. Most notable is the utilization of bedrock as a feature on which architecture could be constructed on a greater scale rather than the smaller architecture at the base of bedrock demonstrated by Phases I and II. Throughout the duration of Phase III, with its many rebuilds of the same general layout, Metallic Ware begins to appear, which may be a later chronological indicator. Phase IV was greatly disturbed, but seems to have retained the orientation of Phase III, if not many of the same lines of previous walls. Therefore, over time there may have been a general tendency for constructional phases to extend further upslope and on top of the natural bedrock, eventually abandoning settlement at the base of the slope.























Geomorphological Context

Reconstructing the depositional sequence in in the vicinity of Megiddo, particularly that in Areas A and B and the JCB cores, necessarily relies heavily upon additional evidence regarding the geology and paleoenvironment in the Jezreel Valley and Tel Megiddo environs. Much has been written about the so-called “Mid-Holocene climatic optimum” that existed from the Chalcolithic period through EB II/III in the southern Levant (Rosen 1986; 1995; 1997; Issar et al. 1992). Evidence for this generally moist episode derives from a number of sources, including isotope analyses of cave speleothems and snail shells (Bar-Matthews et al. 1998; 1999), geomorphology (Goldberg 1994; Rosen 1997), and pollen (Baruch and Bottema 1999). These data indicate moist conditions, significantly wetter than today, from the Chalcolithic period through the start of the EB. During the EB I, the environment remained relatively damp, but with much higher fluctuation in rainfall amount, with the wettest episodes coinciding roughly with the EB I-II transition around 3000 BCE (Rosen 2006, 464). After the drying episode beginning ca. 2200 BCE, the most notable alluvial deposition is the post-Roman/Byzantine-period valley fills, which are ubiquitous throughout the eastern Mediterranean region; in fact, no valley or plain in the southern Levant remained unaffected by this massive episode of alluviation (Rosen 2006, 465; 1986; Goldberg 1994). The derivation of this latter fill may be interpreted as a result of wetter climatic conditions and/or massive deforestation and extensive agricultural activity on the part of the large populations inhabiting the region during the Roman-Byzantine period (Rosen 2006, 465; cf. Goldberg and Bar-Yosef 1990; Bruins 1990). Using this information as a contextual backdrop, the most useful data for reconstructing the depositional sequence of sediment, specifically in Area B, derive from geomorphological observations in two related contexts: (1) Rosen’s 2006 geological survey of the valley within the vicinity of Tel Megiddo, and (2) the 2010 JVRP excavation of unit BB-104 at Tel Megiddo East.

Rosen’s geological survey of the valley surrounding Tel Megiddo consisted of five backhoe trenches that provided geological sections (GS-1-5; Figs. 12-13) allowing an examination of sediment accumulation from the present as far back as the Pleistocene. From these sections, Rosen (ibid., 465; 468) determined that the uppermost “Roman-Byzantine” (or later) fill, Unit 1, existed to a depth of 1-2 m from the surface, described as follows:

It is characterized by heavy black silty clay containing numerous Roman/Byzantine sherds and small angular chalk pebbles. The dark grey color and clayey texture indicate that during the Byzantine period the land around the site of Megiddo was marshy and suffered from poor drainage even upslope from the present-day valley floor. The presence of the angular chalk pebbles indicates that there was a component of colluvial slope wash, suggesting that sheet wash rather than channeled flow was a major form of drainage. This sediment unit suggests that there were high water tables plus much slope wash during deposition in late antiquity.

Unit 2, which extended 2.5 or 3.0 m below the surface, is characterized by heavy dark clays with sometimes dense inclusions of calcium carbonate nodules. The base of Unit 2 (in GS-l, GS-2, GS-3, and GS-5) contained a sub-unit of “small gravels in a clayey matrix . . . representing a network of small streams indicating that runoff was channeled into many small, slow-moving low-energy streams that meandered across the backswamps of the plain” (ibid.). Rosen’s dates for this unit were based on radiocarbon from humic content in the sediments from GS-2 and GS-5, and were respectively 4050-3950 cal. BCE (5200 +/- 60 bp) and 3640-3510 cal. BCE (4740 +/- 45 bp), placing Unit 2 in the Chalcolithic/EB I period. Erosional contact was determined between Units 2 and 3, indicating a long-term break in deposition and a dry erosional episode during that interim. Unit 3 (in GS-2, GS-4, and GS-5) consisted of a light to dark brown clay with heavy carbonate nodule development in GS-2 and higher-energy stream deposits in GS-5 (ibid., 469). The color and development of carbonate nodules led Rosen to interpret this deposit as belonging to the Pleistocene, which was corroborated by radiocarbon determinations from GS-4 with a date of 11,200- 11,010 cal. BCE (11030 +/- 60 bp). Unit 3 was probably formed during the late Pleistocene period of warm and wet conditions following the Late Glacial Maximum (18,000 uncal. bp). Consequently, the erosional contact between Units 2 and 3 probably represents the Younger Dryas (ca. 10,500 cal. BCE), an episode of very cold and dry conditions occurring abruptly between the end of the Pleistocene and beginning of the Holocene, both of which were warmer and wetter. Based on the nature and sequence of units identified in the geological sections, Rosen’s suggested reconstruction of landscape development in the Megiddo vicinity is as follows (ibid.):

The late Pleistocene landscape consisted of hillslopes that were covered by a Mediterranean Oak woodland . . . drained by slow to moderate perennial streams in the valley near Megiddo which contributed a well-watered floodplain, seasonal marshes and some stable land surfaces with dark soil development. With the onset of cold, dry conditions at the time of the terminal Pleistocene Younger Dryas episode . . . alluviation in the stream valleys ceased and an episode of erosion took place which removed the upper layers of the previous Pleistocene deposits. At the beginning of the Holocene, alluviation of the stream valleys resumed under rapidly warming conditions and higher rainfall. There was a resumption of the old drainage pattern consisting of small perennial streams traversing the seasonally marshy deposits in the valley near the mound. The immediate vicinity around the site was characterized by high water tables, rich organic deposits, and small, low-energy muddy streams crossing the landscape. This would have been prime land for agriculture with naturally irrigated fields . . . This valley alluviation ceased sometime in the late Early Bronze Age, probably with the onset of very dry conditions at the beginning of the second millennium BCE. The previously well-watered valley landscape became eroded at some time between the second millennium BCE and the Roman/Byzantine period.


The 2010 excavations of Tel Megiddo East provided useful geological information near the tell that may be compared with Rosen’s results. In every location, sherds were found belonging only to the Roman/Byzantine periods and/or the Chalcolithic/EB I (see above), which supports Rosen’s interpretation that these are the two periods during which the uppermost fills (Units 1 and 2) were deposited in the valley. Unit BB-104 was situated ca. 400 m to the east Area B and about 60 m north of Area C, in a relatively low-lying agricultural field on the eastern side of the wadi runoff from Ain el-Qubbi (Fig. 2). A probe in BB-104 reached bedrock 2.4-2.9 m below the surface, above which were (1) a ca. 1 m-thick layer of topsoil blending continuously with a layer of dark brown silty clay with R/B and EB I sherds, comparable in every way with Rosen’s Unit 1, (2) a layer of EB Ib occupational debris, (3) a thin layer of coarse pebbles and Chalcolithic/EB I sherds, and (4) a thick (2 m at its thickest extent) layer of very dense clay, of dark grayish-brown color, comparable with Rosen’s Unit 2. A sample of this clay was analyzed using grain-size analysis, determining it to consist of roughly 59% clay, 22% silt and 19% sand, classifying it as actual clay (see Fig. 14).

This clay layer in unit BB-104 was completely sterile, which led to the interpretation that it represents a late Pleistocene or post-Younger Dryas deposit. Subsequently cutting through this clay was a layer of coarse pebbles, in which many worn sherds dating to the Chalcolithic and EB I were intermixed. Furthermore, this pebble layer sloped fairly dramatically downward from east to west and slightly from north to south, demonstrating a dip in the middle that was also quite clear in the west section. The lowest part of the cut into the underlying clay appeared in the westernmost portion of the trench, and in this lowest extent, immediately above the clay was a layer of mixed silty material. All of these observations led to the interpretation of the pebbles as evidence of a relatively high-energy water channel that cut through the earlier clay bed, with the silty accumulation probably representing the eventual slow in-filling of the bottom of this channel during a subsequent period of lower-energy drainage. This drainage must have occurred during the EB I, due to the sherds intermixed in the deposit, but ended prior to the end of the EB Ib, since it was sealed by in situ EB Ib occupation (Fig. 14).


Based on the geomorphological sequence in the valley brought to light by Rosen’s work and the 2010 season at Tel Megiddo East, it is possible to reconstruct the depositional sequence in Area B and provide some interpretations regarding the EB I environment of Tel Megiddo and its immediate environs. It appears that the underlying material encountered throughout the area, the very dense dark greyish brown clay, corresponds with the exact same clay at the bottom of BB-104 and Rosen’s Units 2 and possibly 3. The bottom of the probe in unit D-10 encountered more than a meter of clay, with the lowest material becoming slightly more yellowish brown with many carbonate nodules, which corresponds with Rosen’s Unit 3 in GS-2 and GS-5. The clay deposit was encountered in every square throughout the area, sometimes preserved as high as the topsoil/plow zone. Whereas much of the upper layers of the Pleistocene deposits were probably removed by erosion during the Younger Dryas (Rosen 2006, 469), the “mound” on which Area B is located probably represents a particularly well-preserved portion of this deposit. The layer of pebbles cut into this clay in a number of places (see above), and was covered by later, probably Roman-Byzantine, fill. The pebble layer corresponds to the very same phenomenon that sealed the clay in BB-104, and the base of Rosen’s Unit 2, which she interpreted as representing low-energy, muddy stream channels.

The extent of excavation in Area B allows for further observations to be made regarding the pebble layer: (1) the coarse pebbles (ca. 5 – 12 cm) were angular to sub-angular, moderately- to well-sorted and were generally in a dark grey, silty clay matrix; (2) there appeared to be multiple layers of these pebbles, between which was a lighter brown silty material; (3) intermixed with the pebbles were many sherds dating to the Chalcolithic and EB I periods, having some stratification apparent with the lowest pebble layers containing higher quantities of Chalcolithic sherds; (4) in many cases the sides of these pebble deposits contained lines of larger stones (ca. 30-40 cm), which served as boundaries; (5) underlying the lowest layer of pebbles tended to be a layer of light silty material; (6) the pebbles appeared to make a somewhat curved cut into the clay, as was evident in some sections (see Fig. 5).

The combined observations regarding the pebble layers lead to the interpretation of the pebbles as representing stream channels that cut into the earlier clay deposits and meandered through Area B. The energy flow of these streams was not consistent, and at some points they became silted up, resulting in the layers of pebbles and silt. It also appears that larger stones were distributed along the outside curve of these channels, where the energy would have been highest. The layers of well-sorted, angular pebbles probably represent higher-energy flows of water near the source of runoff; likewise, the silty layers represent lower-energy flows that caused the channels to be periodically in-filled. Due to the sorting of sediments, amount of pottery sherds, topographic situation, and proximity to the Tel Megiddo, it seems most likely that these water channels were caused by heavy to moderate rainfall runoff from the eastern slope of the tel, which is the only area that is naturally topographically higher than the elevated mound on which Area B was located. The water would have cut channels through the underlying clay with fairly high energy, as seen especially by a deep and narrow cut in unit D-10, which appears to be along the edge of the clay mound. Based on the types of sherds collected from the pebble layers, this alluvial activity seems to have been steady throughout the EB I (and possibly the Chalcolithic period) and caused by sustained heavy rainfall rather than hillslope erosion (cf. Rosen 1986; 1997; 2006; Goldberg 1994). This depositional pattern ended just prior to the end of the EB I, as suggested by the EB Ib occupation sealing the stream channel in unit BB-104 and the fact that there were absolutely no sherds found in these deposits dating later than the EB I. Therefore, it seems probable that either rainfall lessened in the region, or human factors affecting drainage patterns occurred during the major shift in settlement from EB I to II.


The Early Bronze Age Settlement: Preliminary Interpretations

One of the most significant accomplishments of the 2011 season was the clear identification of the Early Bronze I settlement. In addition to the plan presented in Figs. 9 and 10, several preliminary observations can already be made:

The earliest cultural material that appeared in Area C was at the bottom of the slope on the western end. The pottery from this earliest phase consists of typical EB I pottery with a preponderance of bow-rims and Grey Burnish Ware that suggests a date in the middle of the EB I (aka EB I A/B). This period of occupation appears to have been restricted to the area to the south of the protruding bedrock which makes up the eastern part of Area C. The bedrock is notably uneven, with many outcroppings and caverns.

With the latest two phases of construction, this settlement expanded eastward onto the bedrock (Figs. 9-10; Units F, G, H). This is significant because it represents a willingness (or necessity) on the part of the builders to engineer their buildings to accommodate the very rough bedrock. This was accomplished through the construction of terrace walls to level out spaces for new construction and/or domestic activity. While no in situ occupational debris was excavated from these latest phases, nothing later than the very late EB I was found. The fills behind the terrace are instructive for dating. They contain late EB I pottery, with a notable lack of Grey Burnished Ware. They also contain fragments of combed metallic ware pottery. The general assemblage is consistent with the late EB I pottery found on the tel in Level J-4, the period of the Great Temple. Also notable in the phases which begin the expansion onto the bedrock is that the walls of this phase are notably of higher quality and of a consistent width of 1.05 m. The thickness of these walls represents two 52.5 cm cubits, which were also used as a standard of construction for the Great Temple (Adams, Finkelstein, and Ussishkin forthcoming).

The overall picture that has emerged from the 2011 season is that the settlement at Tel Megiddo East represents at least the period from the middle of the EB I (Chalcolithic ceramics are present, but no stratified remains have been identified yet) through the very terminal phase of the EB I. According to the comparative ceramic evidence, this settlement was abandoned at the same time as the Great Temple on the acropolis. The team’s preliminary conclusion is that the expansion of the settlement onto the bedrock and the standardization of measurements in architecture are related to the dramatic social changes which facilitated the construction of the Great Temple.

Also of note is the presence of 5 Egyptian sherds from the excavation of Area C, including the base of an Egyptian cylindrical jar made of marl clay. Good examples of similar vessel have been discovered at Tell Abu-al Kharaz (Fischer 2002: 325). These sherds from Area C are some of the northern-most Egyptian material found in the Levant dating to the EB I.

The paleoenvironmental and archaeological data have provided a rough picture of the settlement landscape of the EB I. It appears that the cultic acropolis of the Great Temple stood alone with little or no settlement contiguous with it (future excavations in Area S of the Tel Aviv University Megiddo Expedition will test this assumption). Instead, the settlement that served the Great Temple was located to the east across the wadi (Fig. 11). This is, so far, a distinctive type of relationship between a settlement and a sacred space for this region at this time.

As noted above, previous work in Israel has shown that there were major wet climatic episodes throughout the Chalcolithic period and the EB. The combined evidence from Area B, Rosen’s 2006 study, and unit BB-104 indicates that during the Chalcolithic and EB I in and around Tel Megiddo, the cultic acropolis was surrounded by small meandering streams and backwater marshes. Furthermore, the deposits from this period also support the view that climatic conditions were quite moist, and that hillslope soils must have been stabilized by natural vegetation, since there is little evidence for significant hillslope erosion. The dark-colored clay deposits demonstrate that valley in this vicinity consisted of rich and well-watered farmland that could have hosted sizeable EB I settlements. The streambeds themselves carry evidence of significant EB I occupation nearby, such as the large amount of sherds found amongst the pebbles in Area B and unit BB-104; the respective origins of the sherds in these two areas are the mound of Tel Megiddo and the settlement represented by Area C. These two pockets of EB I occupation probably belonged to a number of settlements built around outcrops of bedrock within close proximity to the tel and its cultic complex.

 Area B established that the space between the eastern slope of Tel Megiddo (Megiddo “Stages”) and the Wadi ‘Ain el-Qubbi was uninhabited. Characterized today by a rounded mound just at the base of the tel, it is surrounded by lower-lying land descending to the north, south, and especially to the east to the bed of the wadi. The mound itself is an outcropping of Pleistocene clay, eroded to roughly its present shape during the Younger Dryas Period, probably as a result of intermittent spring activity. In Area A, just off the mound to the north, this erosional activity is apparent in the severely truncated Pleistocene clay deposit – surviving only around 50 cm above the earlier Miocene Tayiqe marl. The upper centimeters of this eroded clay show very dense calcareous nodule development, demonstrating that the end of this erosional period involved significant drying - the drought of the Younger Dryas (ca. 10,000 BCE). In the aftermath of this dry period, the new topography consisted of a broad area filled with runoff from the spring. Conditions between the mound and the present wadi were swampy and a significant amount of clay was deposited.

By the EB I, the Megiddo region was ideal for agriculturalists, but the space between the tel and the wadi remained uninhabitable (Areas A and B): the low-lying area was marshy, but the mound itself was inhospitable to human settlement. During the Chalcolithic and EB I, a wetter period ensued. Rainfall was significantly higher than the preceding and the succeeding periods. The Area B mound was subject to erosion due to its unique position at the base of the cultic acropolis. Rainfall over most of the eastern side of the tel was channeled directly to the vicinity of this mound, where it carved run-off channels in which first chalcolithic pottery and then EB I pottery, bone, and other cultural debris was deposited. As noted above, pottery from these channels dates exclusively to the Chalcolithic and EB I, and, in at least one channel, appears to be chronologically stratified. A similar chronological pattern emerges from unit BB-104, on the other side of the wadi to the east. Here, Chalcolithic pottery was deposited in a wadi channel, which was subsequently built over by an EB I house.

The direction of flow from the Area B channels appears to be from the tel, and the pottery sherds are not very worn. The pottery accumulation within these channels ended in the Early Bronze Age Ib. Therefore, coinciding with the end of the EB Ib, a significant change in the local hydrology occurred, such that run-off that had previously cut channels and deposited pottery ceased. Two options present themselves to explain this change. The first is topographical. In the late EB Ib, the construction of the monumental Great Temple (Level J-4) on the tel changed the topography of the acropolis with the creation of large terraces, including a massive 4 m-wide wall about 50 m upslope from the location of these channels (Adams, Finkelstein, Ussishkin forthcoming). No doubt the overall hydrological situation would have been affected, but it is not clear whether the terracing and other construction specifically impacted the erosion at Area B. The second option is that the water channels became defunct with the changes in weather patterns that accompany the development of the EB (see above). In this case, this transition to a dryer climate would coincide with the abandonment of the Great Temple and the Area C settlement. It is still too early to tell if the data supports a cause-and-effect relationship between these two phenomena.


Hellenistic, Roman, and Byzantine Periods: Preliminary Interpretations

Remains from the Classical period are primarily surface finds from informal survey and from the topsoil removed during excavation at Tel Megiddo East and ‘Ain el-Qubbi. Both the informal survey and the topsoil produced 10-20% Roman/Byzantine pottery, 70-80% Early Bronze I pottery, and ca. 5% Chalcolithic pottery. Nearly all of these topsoil finds were relatively small and well-worn. The dark, silty topsoil ranged in depth from 20 to 80 cm below the surface, depending on the excavation area, and the layer appears to indicate the modern agricultural plow zone. For the most part, these topsoil finds did not reflect the subterranean in situ archaeological remains. In fact, they seem to suggest the introduction of soils from other locations to the farmed areas of Tel Megiddo East and ‘Ain el-Qubbi, whether during modern agricultural activities or at some stage in post-Roman antiquity. Overall, these observations highlight the problems with surface sherd survey as a means for identifying archaeological sites, particularly in the Jezreel Valley and other intensely cultivated areas.

The 2010 trial excavations near the height of Tel Megiddo East discovered that the topsoil overlaid a limestone quarry. The quarry blocks are similar in size and character to those found on the eastern slope of Tel Megiddo. The shape and size of the blocks suggest Iron Age or later quarrying. No securely datable remains were found in association with the quarry. The contours of the quarry marks are not dissimilar from the general shapes of blocks evident in the Roman aqueduct system originally documented by Schumacher and later re-examined by Tsuk, and more recently in 2010 by the JVRP (see below), though this quarry’s limestone was generally poorer in quality.

Schumacher’s excavations at the spring identified a feeder channel which brought water from the direction of the center of the tel to the “spring” (Schumacher 1908: 161; 162, Ab. 233).  In his day it was still pouring forth water, but the origin of this water is unclear. Schumacher also identified a water channel parallel to the main road at the base of the eastern slope of the tel (Schmacher 1908: Pl. 1). The University of Chicago also saw this channel, adding it to their plans, but not describing it (Locus 900; Lamon 1935: Fig. 1; Guy 1938: pl. 1-2), but apparently believing it to be of Roman date. This bedrock-cut channel was investigated by Tsuk in the 1980s (Tsuk 1988-89; 2002) and determined to be of the Roman period (ca. 2nd C. BCE) on the basis of technology and hypothesized to be part of the aqueduct system supplying the fort of the VIth Legion Ferrata at Legio (Dahar ed-Dar; Tsuk 2002: 410, Fig. 1). Tsuk identified portions of the channel carved from the bedrock, as well as others which had stone-troughs inserted into larger carved and/or natural channels (Tsuk 2002: Fig. 2). These stone-trough blocks are U-shaped in section and deployed with the open portion of the U facing up, as expected since these form the water channel (Tsuk 2002: Fig. 2).

The 2010 trial excavations at ‘Ain el-Qubbi uncovered a portion of a stone aqueduct leading southeast from the ‘Ain el-Qubbi spring. The team revealed approximately 4 meters of the aqueduct, which was cut down into sterile soil and was intended to have been mostly subterranean. The construction is shoddy, consisting of reused blocks from the earlier Roman system. At least one of the blocks, which would have been a trough in its original aqueduct, was used as a capping stone in this one. No securely datable remains were retrieved from the within the aqueduct. The blocks used in its rebuilding are similar to the U-shaped trough blocks found in the Roman aqueduct system reviewed above. Considering the fact that this run is made of re-used blocks from an apparently better-made aqueduct, this construction must date to the post-Roman period. Its general direction from the spring leads not toward Roman Legio, but rather toward the ruins identified by Schumacher as “Chirbet ‘ain el-Kubbi” (Schumacher 1908, Pl. 1). Future excavation among these ruins may shed light on the date of this re-use.

Finally, some negative evidence from our JCB survey adds a few points concerning the Classical period. In three JCB trenches in the fields to the north and northwest of Tel Megiddo, no archaeological remains, Roman or otherwise, were uncovered. Instead, Pleistocene clay was reached near the surface. In one trench east of the tel, parallel to the wadi which drains the ‘Ain el-Qubbi spring, we excavated down 4.5 meters through colluvium and alluvium. One coin was discovered here, at a depth of approximately 4 meters below the surface, beneath a substantial layer of largely sterile colluvium. This was a civic issue of Akko-Ptolemais, circa 169-164 BCE. The discovery of the coin at this depth adjacent to the spring water-course suggests that the wadi was significantly deeper in the pre-Roman periods. This would mean that in the Bronze and Iron Ages, the spring run-off was quite heavy and gouged a very deep channel. This remained true into the Hellenistic period when the coin was deposited. Subsequently, the channel filled in with colluvium and alluvium. This deposition is possibly tied to the construction of the Roman aqueduct system, which fed the camp of the VIth Legion and the growing settlement of Legio, diverting water from the spring and making the wadi’s flow weak enough for the channel to be filled in. It may also be the case that the spring itself discharged significantly less beginning in the Roman period as deforestation in the area significantly affected the refilling of the aquifers. Future geoarchaeological work at Tel Megiddo East will aim to understand better this hydrological change over time.


Numismatic Finds

The 2011 excavation season produced a total of twenty coins or coin-like objects, of which sixteen could be identified and dated with some degree of certainty. The topsoil layers of areas A and C account for five of these, situated well out of context and well within the recent plow zone. These were a British Mandate penny (c. 1935 CE), a pierced, disc-shaped ornament (perhaps 1800s CE), a Mamluk mint (1250-1517 CE), and a Byzantine coin of Constans II (641-643 CE) from Area A, while Area C produced a Late Roman issue (351-361 CE) and an Antioch mint of Seleucus IV (187-175 BCE) from the topsoil. The more secure strata of Area C produced no further coins, as these layers clearly consisted of the Early Bronze settlement phases.

One coin was also discovered within the JCB test trench that excavated a portion of the modern ‘Ain el-Qubbi wadi channel, at a depth of approximately 4 meters below the surface, beneath a substantial layer of largely sterile colluvium. This was a civic issue of Akko-Ptolemais, circa 169-164 BCE. See the section above for further discussion of the implications of this singular find.

The remaining eight, in addition to three damaged/unidentifiable coins, all came from two adjacent units in Area A on the western side of the field, closest to the (Roman and modern) road. With one exception (a Seleucid issue), these were all Late Roman, dating primarily to the second half of the fourth century and the early decades of the fifth century CE. One third-century CE provincial issue was also found. Coming from scattered contexts, all of these eleven coins were discovered in and around a cluster of stones, some of which had been altered for architectural use, but clearly now were situated in a secondary context. Apparently the result of field-clearing activities, this stone pile/pit may be dated to the Late Roman period (early 400s CE) on the basis of both the fragmentary ceramic remains and the date range of these coins.



The 2010 and 2011 excavation seasons of the Jezreel Valley Regional Project at ‘Ain el-Qubbi, Tel Megiddo East, and Legio were a great success. The team accomplished its primary goal of identifying the location of the Early Bronze Age I settlement associated with the cultic acropolis of the Great Temple. These excavations have shown that this settlement is not contiguous with the cultic acropolis, as previously assumed, and the arrangement represents a unique settlement pattern for the period. Within the settlement we have documented a significant chrono-cultural transition dating to the late EB I, coinciding with the construction of the Great Temple – the settlement expands, using significant effort and engineering techniques to exploit difficult topography. We have also found that the settlement was abandoned at the terminal EB I at the same time as the Great Temple. Our paleoenvironmental work has fleshed out the ancient topography of the EB I and provided a number of clues that support previous theses that the Chalcolithic and EB I were notably wetter periods, and that the florescence of EB I Megiddo corresponds to this period of wetter climate.

 On a methodological level, the JVRP has provided a valuable example of the dramatic errors that may result from reliance on surface sherd survey to determine the location and size of archaeological sites, particularly in a complex situation like that at Megiddo, where significant ancient and modern agricultural activity has disturbed the topsoil. This study has shown that the plow zone and even deeper levels can be filled with pottery that does not accurately reflect what is below, and that surface sherd survey does not provide and accurate index of archaeological remains. This poses significant difficulties for studies of settlement patterns and settlement hierarchy based on such surveys. At present, our current understanding of human activity in the Jezreel Valley from the Paleolithic to Medieval times is based primarily upon just this type of data recovery, and it therefore must be considered significantly flawed. A better approach to surveying and documenting site size and chronology must be developed, and a new survey conducted in the Jezreel Valley. This is the future task of the Jezreel Valley Regional Project



Adams, M.J. 2011. “A Brief Note on the Architectural Setting of the Great Temple of Megiddo.” Jezreel Valley Regional Project Online Publication Series, 1. (30 May 2011).

Adams, M.J. Forthcoming a. “Area J (The 2004–2008 Seasons). Part III: The Early Bronze Age, Stratigraphy and Architecture.” In Megiddo V: The 2004–2008 Seasons, edited by I. Finkelstein, D. Ussishkin, E. Cline. Monograph Series of the Institute of Archaeology of Tel Aviv University. Tel Aviv: Institute of Archaeology.

Adams, M.J. Forthcoming b. “The Early Bronze Pottery from Area J,” in Megiddo V: The 2004–2008 Seasons, edited by I. Finkelstein, D. Ussishkin, E. Cline. Monograph Series of the Institute of Archaeology of Tel Aviv University. Tel Aviv: Institute of Archaeology.

Adams, M.J. Forthcoming c. “The Egyptianizing Pottery from Megiddo Area J, Revisited. Stratigraphy, Form, Function and Implications for the Three Temples in Antis.”

Adams, M.J. Forthcoming d. “Appendix: Area J The 2010 Season,” in Megiddo V: The 2004–2008 Seasons, edited by I. Finkelstein, D. Ussishkin, E. Cline. Monograph Series of the Institute of Archaeology of Tel Aviv University. Tel Aviv: Institute of Archaeology.

Adams, M.J. Forthcoming. “Area J (The 2004-2008 Seasons). The Early Bronze Age, Stratigraphy and Architecture,” in I. Finkelstein, D. Ussishkin, M. Adams, E. Arie, E. Cline, N. Franklin, and M. Martin (eds.), Megiddo V: The 2004-2008 Seasons (Tel Aviv: Institute of Archaeology, Tel Aviv University).

Adams, M.J., I. Finkelstein, and D. Ussishkin, The Great Temple of Early Bronze Age Megiddo

Bar-Matthews, M., Ayalon, A. and Kaufman, A. 1998. Middle to Late Holocene (6,500 yr. period) paleoclimate in the eastern Mediterranean region from stable isotopic composition of speleothems from Soreq Cave, Israel. In: Issar, A., Brown, N. and Shilony, Z., eds. Water, Environment and Society in Times of Climatic Change. New York. pp.203-2l4.

Bar-Matthews, M., Ayalon, A., Kaufman, A. and Wasserburg, G.J. 1999. The eastern Mediterranean paleoclimate as a reflection of regional events: Soreq Cave, Israel. Earth and Planetary  Science Letters 166(1-2):85-95.

Baruch, U. and Bottema, 5. 1999. A new pollen diagram from Lake Hula: vegetational, climatic and anthropogenic implications. In: Kawanabe, H., Coulter, G.W. and Roosevelt, A.C., eds.  Ancient Lakes: Their Cultural and Biological Diversity. Brussels. pp. 75-86.

Briken, Andreas. 1976. Die Provinzen des Osmanischen Reiches (Beihefte zum Tübinger Atlas des Vorderen Orients, Reihe B (Geisteswissenschaften) Nr. 13; Wiesbaden: Dr. Ludwig Reichert).

Bruins, H.J. 1990. The impact of man and climate on the central Negev and northeastern Sinai deserts during the Late Holocene. In: Bottema, S., Entjes-Neiborg, G. and Van Zeist, W., eds. Man ‘s Role in the Shaping of the Eastern Mediterranean Landscape. Rotterdam. pp. 87-99.

Eppelbaum, L., and S. Itkis. 2000. “Magnetic Investigations in the Proto-Historic Site to the East of Tel Megiddo,” in I. Finkelstein, D. Ussishkin, and B. Halpern (eds.), Megiddo III: The 1992-1996 Seasons (Tel Aviv: Institute of Archaeology, Tel Aviv University) 504-513.

Engberg and Shipton. 1934.

Finkelstein, I., B. Halpern, G. Lehmann, and H. M. Niemann. 2006. “The Megiddo Hinterland Project,” in I. Finkelstein, D. Ussishkin, and B. Halpern (eds.), Megiddo IV: The 1998-2000 Seasons (Tel Aviv: Institute of Archaeology, Tel Aviv University) 705-776.

Finkelstein, I., and D. Ussishkin. 2000. “Area J,” in I. Finkelstein, D. Ussishkin, and B. Halpern (eds.), Megiddo III: The 1992–1996 Seasons (Tel Aviv: Institute of Archaeology, Tel Aviv) 25-74.

Fischer, P.M. “Egyptian-Transjordanian Interaction during Predynastic and Protodynastic Times: The Evidence from Tell Abu al-Kharaz, Jordan Valley,” in van den Brink, E.C.M., and T.E. Levy (eds.) Egypt and the Levant. Interrelations from the 4th through the Early 3rd Millennium B.C.E. London: Leicester University Press. 2002. 323-333.

Goldberg, P. 1994. Interpreting Late Quaternary continental sequences in Israel. In: Bar-Yosef, O. and Kra, R.S., eds. Late Quaternary Chronology and Paleoclimates of the Eastern Mediterranean. Tucson. pp. 89-102.

Goldberg, P. and Bar-Yosef, 0. 1990. The effect of man on geomorphological processes based upon evidence from the Levant and adjacent areas. In: Bottema, S., Entjes-Neiborg, G. and Van Zeist, W., eds. Man ‘s Role in the Shaping of the Eastern Mediterranean Landscape. Rotterdam. pp. 71-85.

Greenberg, R. 2002. Early Urbanizations in the Levant, A Regional Narrative (London: Leicester University Press).

Issar, A.S., Govrin, Y., Geyh, MA., Wakshal, E. and Wolf, M. 1992. Climate changes during the  Upper Holocene in Israel. Israel Journal of Earth Sciences 40:219-223.

Raban, A. 1999. Archaeological Survey of Israel: Map of Mishmar Ha’emeq (32) (Jerusalem: Israel Antiquities Authority).

Rosen, A.M. 1986. Quaternary Alluvial Stratigraphy of the Shephela and Its Paleoclimatic Implications. (Geological Survey of Israel Report GSI/25!86) Jerusalem.

Rosen, A.M. 1995. The social response to environmental change in Early Bronze Age Canaan. Journal of Anthropological Archaeology 14:26-44.

Rosen, A.M. 1997. Environmental change and human adaptational failure at the end of the Early  Bronze Age in the southern Levant. In: Dalfes, H.N., Kukia, G. and Weiss, H., eds. Third  Millennium BC Climate Change and Old World Collapse. Berlin. pp. 25-38.

Rosen, A. M., 2006. Climate Change, Landscape and Shifting Agricultural Potential. In I. Finkelstein,  D. Ussishkin and B. Halpern (eds) Megiddo IV: The 1998-2002 Seasons. Tel Aviv: Emery  Claire Yass Publications in Archaeology. 463-70.

Rosen, A. M. 2007. Civilizing Climate: Adapting to Climate Change in the Ancient Near East, from Foraging Societies to Empires. Lanham, MD.

Schumacher, Gottlieb. 1908. Tell el-Mutesellim 1 (Leipzig).

Tepper, Y. 2002. Lajjun – Legio in Israel: Results of a survey in and around the military camp area. In Freeman, P. Bennett, J. Fiema. Z.T. and Hoffmann, B. Eds. Proceedings of the XVIIIth International congress of Roman frontier studies, Amman, September 2000.Limes XVII BAR I.S. 1084 (1): 231-242.

Tepper Y. 2007. The Roman Legionary Camp at Legio, Israel: Results of an Archaeological Survey and Observations on the Roman Military Presence at the Site. In Lewin, A.S. and Pellegrini, P. eds. The Late Roman Army in the Near East from Diocletian to the Arab Conquest. BAR International Series 717, Oxford: 57-71.

Tsuk, Tsvika. 1988-89. “The Aqueduct to Legio and the Location of the Camp of the VIth Roman Legion,” Tel Aviv 15-16: 92-97.

Tsuk, Tsvika. 2002. “An Aqueduct to Legio,” in David Amit, Joseph Patrich, and Yizhar Hirschfeld (eds.), The Aqueducts of Israel (JRA Supp. 46; JRA: Portsmouth, Rhode Island) 409-411.

Figure 1. Tel Megiddo East, showing the locations of Areas A, B, and C, excavated in the 2011 season. Image via Google Earth.

Figure 2. Topographic plan of Tel Megiddo East, showing Areas A, B, and C of the 2011 Season. Note also the location of the 2010 GPR grid, in which Areas A and B are situated.

Figure 3. Greater Megiddo area, showing the locations of mechanical backhoe trenches conducted as part of the paleoenvironmental study. Points are numbered 1 through 7 from east to west.

Figure 4. Anomaly from Area B.

Figure 5. Ortho-rectified section through one of the Area B anomalies.

Figure 6. Excavation Areas A, B, and C. Green represents Area B anomalies as reconstructed water channels. Image via Google Earth.

Figure 7. One sondage in Area A, showing lacustrian clay deposits to a depth of 2 m.

Figure 8. Excavation Areas A, B, and C. Blue represents reconstructed pre-Roman water course from the spring. Image via Goggle Earth.

Figure 9. Combined top plan of EB I architecture from Area C.

  Figure 10. Photogrammetric plan of Area C.

Figure 11. Schematic reconstruction of the relationship between the cultic acropolis of the Great Temple and the EB I settlement.

Figure 12. Location of geological sections cut by backhoe (after Rosen 2006, Fig. 21.1).

Figure 13. Schematic drawings of Geological Sections 1, 2, 4, and 5 (after Rosen 2006, Fig. 21.2).

 Figure 14. Schematic drawings of BB-104, west and north sections.


Copyright Jezreel Valley Regional Project © 2016

All essays appearing on this website are authored by members of the JVRP.

Authorship credit is given where appropriate, as is credit for revisions and additions.

When citing from any of this material, please cite the credited authors

and note the date and time retrieved, as all content is subject to update and revision.