Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

First evidence of “ancient deer” (cervid) in the late Miocene Bira Formation, Northern Israel

  • Alexis Gabriel Rozenbaum ,

    Roles Conceptualization, Formal analysis, Investigation, Writing – original draft, Writing – review & editing

    Affiliations Geological Survey of Israel, 30 Malkhe Israel, Jerusalem, Israel, Institute of Earth Sciences, The Hebrew University of Jerusalem, Jerusalem, Israel, Edmond J. Safra Campus, Givat Ram, Jerusalem, Israel

  • Dotan Shaked Gelband,

    Roles Data curation, Investigation, Writing – original draft, Writing – review & editing

    Affiliations Geological Survey of Israel, 30 Malkhe Israel, Jerusalem, Israel, Institute of Earth Sciences, The Hebrew University of Jerusalem, Jerusalem, Israel, Edmond J. Safra Campus, Givat Ram, Jerusalem, Israel

  • Mordechai Stein,

    Roles Conceptualization, Funding acquisition, Writing – original draft, Writing – review & editing

    Affiliations Geological Survey of Israel, 30 Malkhe Israel, Jerusalem, Israel, Institute of Earth Sciences, The Hebrew University of Jerusalem, Jerusalem, Israel, Edmond J. Safra Campus, Givat Ram, Jerusalem, Israel

  • Henk K. Mienis,

    Roles Conceptualization, Data curation, Investigation, Methodology, Validation

    Affiliation Natural History Collections, The Hebrew University of Jerusalem, Edmond J. Safra Campus, Givat Ram, Jerusalem, Israel

  • Rivka Rabinovich

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing

    Affiliations Institute of Earth Sciences, The Hebrew University of Jerusalem, Jerusalem, Israel, Edmond J. Safra Campus, Givat Ram, Jerusalem, Israel, Natural History Collections, The Hebrew University of Jerusalem, Edmond J. Safra Campus, Givat Ram, Jerusalem, Israel

First evidence of “ancient deer” (cervid) in the late Miocene Bira Formation, Northern Israel

  • Alexis Gabriel Rozenbaum, 
  • Dotan Shaked Gelband, 
  • Mordechai Stein, 
  • Henk K. Mienis, 
  • Rivka Rabinovich


Despite the extensive geological and paleontological searches in the south Levant, no terrestrial fauna of late Neogene age was yet reported. Here, we report the first evidence of “ancient deer”–cervid in the late Miocene (Tortonian) lacustrine section of the Bira Formation at Hagal Stream, Jordan Valley, northern Israel. The section comprises rich assemblage of macrofauna fossils, mostly freshwater mollusks. The mammalian bone was discovered among the macrofauna fossils, and is described as an almost complete left humerus of an adult animal identified as an artiodactyls element probably of a cervid. This terrestrial mammal shares similar paleoenvironmental conditions with other contemporaneous localities, where cervids prevailed with bovids and other taxa. It appears that the freshwater lakes provided favorite habitat for the development of the cervids and possibly other mammals. The specific conditions of preservation of the cervid illuminate the role of post depositional processes (taphonomic constrains) in masking terrestrial fauna remains in the region. Accordingly, further efforts will be devoted to unveil the mute part of the southern Levant–east Mediterranean terrestrial faunistic realm at the end of the Miocene.

1. Introduction

The Miocene-Pliocene periods were distinguished by significant tectonic deformation and magmatic activity throughout the Arabian and African plates [13]. The breakup of the Arabian-African continent was accompanied by the opening of the Red Sea, the formation of the Suez rift and the Dead Sea Transform (DST) [4, 5] (Fig 1A and 1B). The rifting process was accompanied by widespread magmatic activity, mainly eruption of alkali basalts [6 and references there]. At the early to middle Miocene (~ 20–10 Ma) the areas that extend now over the Lower Galilee- Yizre’el Valley and the Jordan Valley (Fig 1B) were characterized by a low relief with a large river systems draining from east to west and depositing sequences of fluvial/alluvial sediments (the Hordos Formation, e.g. [79]). During this time, extensive volcanic eruptions occurred over the Arabian plate producing in northern Israel and Golan Heights thick sequences of alkali basalts that comprise the Lower Basalt unit (spanning ~ 17.5–10 Ma -, see Fig 1B) (e.g. [7, 911]). At the same time the morphological basins and valleys of northern Israel deepened (e.g. [5, 7]). This was accompanied by a reversal in the direction of the drainage system towards the east where the Jordan rift valley, Kinnarot and Dead Sea basins developed (e.g. [7, 9]). Water-bodies, mainly freshwater to brackish lakes [12, 13] filled the basins and valleys in northern Israel depositing sequences of sediments that comprise the Bira and Gesher Formations (described below) [7, 14]. The transition from a broad drainage system that flowed over a “prairie”- type landscape to a “basin” and “around lakes” environment had a profound impact on the forthcoming ecological-evolutionary development of the region [15].

Fig 1.

(a) Regional location map; (b) Simplified geological map of northern Israel (modified after [34]); (c) Geological map of the Hagal Stream on the Kinnarot Basin (modified after [17, 35]).

In this paper, we report the first evidence for terrestrial mammal in the late Miocene sedimentary sequence of the Bira Formation at Hagal Stream, located at western margin of the Jordan Valley, in the eastern Lower Galilee area (Fig 1B). The Bira Formation of Tortonian age (~ 10.0–7.0 Ma) [11] comprises a sequence of sediments (limestones and dolomites with marls, gypsums and halite) [16] that were deposited from the lake’s fresh-brackish waters (e.g. [14, 16]). Continental fauna of that period is known from the Mediterranean basin, though hardly from the southern Levant. We describe the deposition environments of the Bira Formation at the Hagal Stream section, present the megafauna fossils from this section and discuss the implications of the discovery of the mammal bone for the evolution of terrestrial late Neogene fauna in the vicinity of Israel.

2. The Tortonian-Messinian periods in the circum-Mediterranean

During the late Miocene–early Pliocene times (~ 10 to 5 Ma) the morphological basins and the lowlands that extended over the areas of the Yizre’el Valley, lower Galilee, Kinnarot Basin and Jordan Valley (Fig 1B) were occupied by lacustrine-estuarine and palustrine water-bodies that deposited sequences of carbonates (limestones and dolomites), marls, gypsum and halite (in the Kinnarot Basin only) (e.g. [7, 14, 1618]). These sedimentary sequences comprise the Bira and Gesher Formations of Tortonian (~10 to ~ 7 Ma) and Messinian-Zanclean (~ 7 to ~5 Ma) times, respectively [11]. During this time-interval marginal lagoons and lacustrine water-bodies developed in the circum-Mediterranean area (e.g. the Pannonian Basin, [1921]). During the Tortonian period (~11.6 to 7.2 Ma) the Mediterranean was connected to the Atlantic Ocean via several gateways off southern Spain and northern Morocco [22]. Later, during the Messinian period (~7.2 to 5.3 Ma) the Mediterranean Sea became progressively isolated from the Atlantic Ocean, generating widespread precipitation of evaporites (~5.9 to 5.6 Ma), massive salt deposition (~5.6 to 5.5 Ma) and a dramatic sea level lowering (~5.5 to 5.3 Ma) during the Messinian Salinity Crisis (MSC) (e.g. [2324]). The MSC was abruptly ended at the early Pliocene (~5.3 Ma) by catastrophic flooding (the Zanclean transgression). The Mediterranean returned to open marine conditions [25]. All these processes had their impact on the environmental history of the circum-Mediterranean areas and the development of the regional fauna before the onset of the tectonic movements that caused regional uplifts in the marginal areas of the Mediterranean and the onset of the Ice Ages in the Pliocene-Pleistocene [15, 2628].

3. The late Miocene Bira and Gesher Formations

A sequence of sedimentary rocks termed the Bira and Gesher Formations was deposited in lacustrine water bodies that occupied during the late Miocene-early Pliocene periods the areas of the Lower Galilee, Yizre’el Valley and the Jordan Valley. The formations were deposited between the time of eruption of the Lower Basalt and the Cover Basalt Units (between ~10 and ~ 5.3 Ma) (e.g. [7, 14]. The depositional regime of the Bira and Gesher Formations was mostly lacustrine. This regime was occasionally interrupted by transgressions of Mediterranean Sea waters (e.g. [7, 14, 16, 29, 30]). Coinciding with the existence of the lakes in the Lower Galilee, Yizre'el and Jordan Valley margins, evaporitic water-bodies were developed in two unique tectonic basins that were associated with the tectonic activity along the Dead Sea Transform Fault [1, 31 and references herein]) and the Kinnarot Basin (Fig 1B and 1C) where the Zemah-1 borehole recovered ~ 4.25 km of carbonates, marls, gypsum, halite and basalts [32] and the Dead Sea Basin, where boreholes recovered several thousand meters of fluvial, lacustrine and evaporitic sediments comprising the Sedom Formation [29]. Thick sequences of salt from marine origin were deposited in the Kinnarot and Dead Sea Basins, respectively (e.g. [29, 3133]).

The Tortonian (~10–7.2 Ma) Bira Formation, where the mammal remains were discovered, comprises limestones, dolomites, marls, gypsum sediments and basaltic rocks (Figs 1C and 2) and halite at the Kinnarot Basin. The lacustrine water bodies of the Bira Formation comprehend shallow water-bodies that extended from the western border of the Lower Galilee in the west to the Golan Heights in the east including the area of the Pleistocene–Modern Sea of Galilee [7, 16]. The surface levels of Lake Bira changed frequently allowing the setting of evaporitic conditions and formation of dolomites [36]. The lacustrine regime was interrupted by two prominent marine (Mediterranean) ingressions, at the base and the top of the Bira Formation (e.g. [7, 14, 16, 30]).

Fig 2. NW-SE cross section A-AA along the Hagal Stream (modified after [17]).

4. The Bira and Gesher Formations at the Hagal Stream section

Here, we focus on the Hagal section of the Bira Formation. The section is located at the eastern side of Hagal Stream, at the western escarpment of the Dead Sea Rift (Fig 1B and 1C). The section forms a recumbent fold, part of a half-dome located between two shear zone faults of the western Jordan Valley margin (Figs 1C and 2). The layers dip sharply (~ 70 to 40°) from SE to NW (Fig 1C and Fig 2). The Bira Formation at the Hagal section comprises about 50 meters of fossiliferous carbonates, which form benches in the landscape, and soft units of gypsum, which are mostly covered by soils or talus (Fig 3). The base of the formation is not exposed. The fossiliferous benches contain gastropods and bivalves within calcarenite matrix with carbonate cement (Fig 4 and see below). Some beds consist of fragments to pulverized fossils, suggesting high water energy conditions. On one of these benches, a mammal bone (humerus) (Figs 3 and 5) and a fish jaw were found. Gypsum beds comprise nodular, thin bedding and ripple marks structures. The contact between the Bira and Gesher Formations is erosive. The base of the Gesher Formation is not exposed at the Hagal section. The lower part of the formation comprises a polimictic conglomerate that is overlain by ooidic limestones, limestone breccia and beds and nodules of chert. The top of the section consists of soft limestones and dolostones (Fig 3).

Fig 3. Lithostratigraphy of the Hagal Stream Section.

The top of the section is at coordinates 35°33ʹ01ʺ, 32°37ʹ57ʺ and elevation 140 meters below sea level. Legend indicates the lithology and main features symbols observed in the columnar section.

Fig 4. The molluscan fossil assemblages of the Hagal section.

(a) Loosely packed moderately fragmented bivalves (mainly Unionidae) in their original shells from the upper part of unit HF1; (b) Densely packed scarcely fragmented Neritidae gastropods (Theodoxus spp.), preserved as external printings or inner molds, comprising the monospecific assemblage of unit HF2; (c) Densely packed extensively fragmented bivalves and gastropods of unit HF3, Dreissenidae mold and Melanopsidae imprint are identifiable; (d) Sparse-loosely packed complete Assimineidae gastropods, comprising the monospecific assemblage of HF4.

Fig 5. The block from Hagal section (unit HF3) in which the bone was found.

The bone was embedded into the sediment. Notice the severe erosion of the shaft and the proximal epiphyses.

5. The macrofaunal assemblage at Hagal section

5.1 General description

The macrofaunal remains at Hagal section include gastropods and bivalves of freshwater origin; some of them tolerate or prefer brackish conditions. Neritidae (Theodoxus spp.) and Unionidae are the most abundant gastropods and bivalves, respectively. The Bira Formation is divided into three units (HF1-3), according to the macrofaunal assemblages described below. Another unit (HF4) appears in the above-lying Gesher Formation (Table 1).


The lowermost beds of the exposed section comprise calcarenite with loose-densely packed shells (following shells close-packing definitions by [37]). The assemblage characterizes by 10 freshwater mollusk families, dominated by gastropods (88% abundance). Neritidae (Theodoxus spp.) is the most abundant taxa, but with relatively low dominance of 47%. Other gastropod families include Cochliopidae, Bithyniidae, Thiaridae, Melanopsidae, and Planorbidae. Bivalves, mainly Unionidae, show increase in abundance towards the upper part of HF1, where they appear moderately fragmented in their original shells (Fig 4A).


This unit comprises brown-reddish soft bio-calcarenite of fine size with 2.5% pyroclasts of silt and sand size and densely packed shells. The assemblage is characterized by very low species richness of two gastropod families, with very high dominance of Neritidae gastropods (Theodoxus spp.), comprising 97% of the assemblage. Most of the fossils in this nearly monospecific assemblage appear as external printings, sometimes with preserved colored sculptures, some appear as inner molds (Fig 4B). Fragmentation degree is low to moderate.


The upper 25 meters of the Bira Formation comprise light gray to light brown soft bio-calcarenite beds that characterized by fine grain size, with 2.5–5% pyroclasts of silt to fine pebble size, and contain mainly crushed shells (Fig 4C). Occasionally, numerous specimens appear unbroken in the surrounding fragments, especially of the epifaunal families Neritidae, Melanopsidae and Dreissenidae (Fig 4C). This could be the result of different shell strengths, sizes, thicknesses and shapes affecting the degree of resistance to breakage [38]), or imply a complex taphonomical history of recolonization by next generations (e.g. [39]). The assemblage of the recovered fossils from the crushed surrounding of unit HF3 is characterized by median species richness of 5 families, with low dominance of the most abundant taxa, which is of Neritidae family, comprising 44% of the assemblage.


A single fossilized bed is identified in the Gesher Formation. The bed consists of a brown highly indurated ooidic limestone, with ooids of fine sand to silt size, and sparse-loosely packed shells. The assemblage is comprised entirely from brackish minute gastropods of the Assimineidae family, comprising 100% of the assemblage (genus Assiminea, Fig 4D).

5.2 Environmental conditions

The paleontological and lithological characteristics of the Bira Formation at the Hagal section (Fig 3 and Table 1) indicate that the water-body was strongly influenced by salinity changes. Variations in the community structure between assemblages, expressed by the ecological parameters of species richness and dominance, indicate variations in the environmental stress levels (e.g. [4042]). Monospecific assemblages are rare and commonly result from opportunistic species taking advantage of unusual environmental conditions that would be stressful for most species (e.g. [40, 4344]). The variations in the species richness (see Table 1), showing a major decrease from the relatively rich assemblage of unit HF1 to the nearly monospecific assemblage of unit HF2, indicates that environmental stress levels increased. Then, the increase in species richness from HF2 to the relatively rich assemblage of HF3, indicates environmental stress levels decreased again. The same pattern is presented by the variations in the abundance of the dominant Neritidae gastropods, from less than 50% of the assemblage in HF1, to complete dominance in HF2, and back to less than 50% of the assemblage in HF3. These shifts in dominance, together with the highly euryhaline nature of these gastropods [4546], strongly suggest salinity was the main factor controlling community structure [40, 47]. The alternating lithology, between carbonatic rocks rich with fresh-brackish fossils and gypsum layers, support this conclusion and further imply the influence of salinity variations in the investigated section, from periods of fresh-brackish water to periods of hypersaline conditions.

The taphonomic characteristics of the macrofaunal assemblages of the Bira Formation in the Hagal section indicates that changes in water depths and energy levels were dominant factors in assemblage formation. The fragment shapes and sizes and the lack of predation traces in unit HF3, indicate the extensive breakage was caused by washing and pounding of the shells on a high-energy shore [38, 48, 49] The sharp taphonomic shift with time from low-medium fragmentation (units HF1, HF2) to massive breakage (unit HF3), implies a shift from relatively deep and quiet conditions to shallower, more energetic conditions very close to the lake’s shorelines. The disappearance of gypsum beds (Fig 3) and the finding of the mammalian bone (see below) in unit HF3 also support the suggested decrease in water depth and setting near the shorelines of the lake.

The complete monospecific character of unit HF4 from the Gesher Formation at the Hagal section imply high levels of environmental stress (e.g., [4342]. In lack of other fossilized bed in the Gesher Formation., it is difficult to fully understand the general environmental conditions during this period. However, the minute gastropods that comprise this entire assemblage, Assiminea sp., is known to inhibit brackish water bodies [5051], therefore the high environmental stress indicated by the monospecific appearance can possibly be related to salinity, similarly to the Bira Formation. The appearance of dolomite layers and disappearance of gypsum layers in the Gesher Formation can point to relatively less hypersaline conditions during this deposition period, compared to the Bira Formation.

5.3 Comparison with other sections in the region

An important conclusion so far is that variations in salinity characterized Lake Bira at the Hagal Stream area, from periods of fresh to brackish conditions during which carbonate rock rich with macrofauna were deposited, to hypersaline-evaporated conditions during which gypsum beds were deposited. Similar conditions also prevailed in other areas, for example the Hurvat Ze’ev section, located east to Hagal section, at the Tabor Stream area (Fig 1B). The macrofaunal assemblage of the Hurvat Ze’ev section is mostly fresh-brackish water fossils, with similar taxonomic family list as the Hagal section. However, the most abundant taxa in these two sections is different, with Thiaridae family (Melanoides spp.) as the dominant taxa in Hurvat Ze’ev, and Neritidae family (Theodoxus spp.) as the most dominant taxa in Hagal section. The more euryhaline and opportunistic nature of the Neriitida family, compared to the Thiaridae family, points to increased salinity-stressed conditions in the Hagal section.

Marine assemblages comprise 10% of the fossiliferous Hurvat Ze’ev section, and are distributed in five beds along the section that represent transgression events of Mediterranean Sea water. These marine assemblages are nearly monospecific, comprising almost entirely from small-sized euryhaline bivalves of the Veneridae family (Venerupis sp. and Paphia sp.), indicating high levels of environmental stress during these saline episodes [52]. Furthermore, the Hurvat Ze’ev section characterizes by appearance of relatively thin dolomite beds, which are related to periods of significant evaporation and interpreted as related to climatic changes toward arid periods [36]. Thus, it seems that the evaporitic conditions were more intense when moving east to the Hagal area, enabling the deposition of gypsum beds. Further to the west, at Migdal Ha’Emeq section of the Bira Formation. (Fig 1B), a dramatic change is observed in the macrofaunal assemblages, reflected in the distribution of fossiliferous beds and the species composition. Only two beds are rich with macrofauna, indicating that for most of the time the conditions on the bottom floor were unsuitable for colonization. These two assemblages are comprised entirely from marine species, and their community structure properties indicate good ecological conditions. The rest of the Migdal Ha’Emeq section is composed of carbonate rocks, mainly limestones. The disappearance of the fresh-brackish community towards the west indicate more saline conditions and the rich marine assemblages imply periods of major sea water intrusions. The connection to the Mediterranean is decreasing to the east, through the ecologically-stressed assemblages in the Hurvat Ze’ev section, and to the entirely fresh-brackish assemblages of the Hagal section. The solution that deposited the gypsum beds in the Hagal section might have originated from these marine transgression events.

5.4 The mammalian bone

In a block that fell from unit HF3 (Fig 3), the outlines of a limb bone were observed. The bone was embedded in the sediment and the exposed part was already eroded when found (Fig 5). It first seemed that the compaction in the crushed mollusks had deformed the general outline. Being such a rare find, apparently the first known mammalian limb bone from the late Miocene in the area, we invested every possible effort to expose the limb. Combinations of diluted acid solution and mechanical cleaning were required in order to finally release the bone from the sediment. However, the complete exposure of the bone outer surface was not entirely possible as the sediment is "imbedded" in the bone (Fig 6A). Microscopic observation supported the necessary final observation for the outer shape of the bone. Black miniature spots probably due to organic contamination were observed over the bone. This possible remnant of organic material is under study (Fig 6B). Taphonomic observations are partially masked by the embedded sediment coverage. The erosion of the shaft is probably due to water wave movement along the shores of the lake.

Fig 6.

Close up of the bone surfaces: (a) sediment compacted still adherent to the bone surface; (b) black spots probably due to organic contamination.

Description of the bone (NGM1): almost complete left humerus of an adult animal (proximal epiphysis fused). The proximal head is not complete and the lateral tuberosity is missing (Fig 7C–7E), the shaft along the lateral view is missing (Fig 5 and Fig 7C), the distal end is almost complete, though parts are missing from the lateral condyle (Fig 7B and 7C). In the absence of the shaft (e.g., missing the area of the deltoid tuberosity) and most of the proximal part (e.g., lateral tuberosity and inter tuberal groove) the possibility to assign it to a species is limited.

Fig 7.

Various aspects of the humerus from Hagal section (unit HF3): (a) medial view; (b) posterior view; (c) lateral view; (d) anterior view; (e) proximal head; (f) distal condyle.

Following the characteristics of the distal part described by [53, Vol. I, Planche XXXVI, 6; Vol. II, p.30], medial valley is situated high in comparison with the external condyle, the external lip is acute and the internal condyle surpasses it, as is typical in cervids. We suggest that this humerus is similar to a cervid. However, we are aware of the problematic of identifying a genus in the absence of teeth and cranial parts.

Family Cervidae Gray, 1821

Cervidae gen. and sp. indet

Humerus (left, NGM1), Hagal Stream, Unit HF3, Bira Formation, late Miocene (Tortonian).

Measurements of the bone (NGM1): Humerus Cervidae gen. and sp. indet

GL (total length on anterior view) - 25.348 mm

Bp (breadth of the proximal end)—ca. 82 mm

BT (breadth of the distal condyle) - 59.85 mm

Bd (breadth of the distal end) - 58.40 mm

SD (smaller breadth of diaphysis)—ca. 34.75 mm

Tortonian localities are rich in terrestrial vertebrate animals from small herbivores to large proboscideans. Tortonian mammals are known from the Kefraya and Zahle Bekaa Valley in Lebanon [5455]. [5657] reported the presence of a proboscidean tooth from sediments of the Gesher Formation at a site located northeast of the Sea of Galilee, near Sheikh Ali [57 p.322].

Fossil cervidae are part of the faunal components of early Miocene onwards in Eurasia. The nearest ancestors of the recent taxa were identified in the late Miocene, though there is a dispute about the relation between the ancestor's origins. Many species were identified based on the antler and dentition [58 and references herein]. Cervids were reported from many late Miocene localities from the Levant and the Mediterranean realm, to mentioned just a few: the areas of Maragheh Iran [59], many sites from Turkey [60] Greece [6162], Italy[63] and Spain [64].

This sole limb is just a testimony of a probable rich fauna invisible due to post depositional processes that buried most of the Tortonian localities in the study area. In the absence of antler and teeth how can we better identify the current element?, and can we speculate more on the basis of the available humerus properties? The length of the element (GL) excludes the presence of any medium cervid as is much larger than the local Dama (Dama mesopotamica) that probably prevailed from the early Middle Pleistocene onward [65]. [66] suggested to examine the mid-diaphysis properties of long bones by looking at the mid-diaphyseal properties of limb long bones, characterizing the thickness of the bone walls, but actually found little correlation between the humeri and the allometry of cervidae species. As yet, a certain thinning of the cortical wall and bending was observed on larger species as can be noticed on the NGM1 humerus (Fig 7).

The Tortonian paleo-environment of local fresh water-bodies offers ambient conditions to terrestrial species such as large size cervids. The paleontological and genetic data refer to the radiation of old world deer [67] during the late Miocene. Earlier evidence points towards eastern Asia as the region of origin of the group sensu lato [68 and references herein]. Geographical buries prevented continues dispersal, though larger cervids with greater territories and flexible dietary needs have advantageous in dispersal. Along the dispersal route from East Asia we can witness sites in Iran [59] crossing through various water-bodies into the studied area.

6. Summary

The recent finding of an artiodactyls element, probably of a cervid, from the late Miocene section of the Bira Formation contributes to the local paleoenvironmental reconstruction. The salinity of the water-body (Lake Bira) was mostly fresh, as indicated by the molluscan identification. A variety of organisms of Tortonian times undoubtedly used this lake as an important water source. The mammalian bone was found in unit HF3, in which a major decrease in water depth and increase in water energy is indicated by the extensive fragmentation and pulverizing of the fossils. The cervid probably died for some reason while drinking water from the lake shorelines. Thus, the finding of the mammalian bone reinforces the reconstruction of the depositional environment as shifting to shorelines conditions. The strong wave action rolled the outer surface of the bone and the changing shore water regime caused the sediment compaction. Further survey in the area might reveal auxiliary exposure richer in vertebrate fauna as expected from Tortonian localities.

The development of fresh water-bodies during the late Miocene in the Levant and circum-Mediterranean area opened the route for the ancient deer expansion westward.


We thank Dr. Amir Sandler and Dr. Ezra Zilberman from the Geological Survey of Israel and Yossi Morag of Gesher Quarries who guided us to the Hagal section. Special thanks to Gali Beiner (ACR), the conservator of The Natural History Collections of the Hebrew University that had speared no effort to expose this unique find and Uzzan who took the photos of the bone. Field assistance by Yaakov Rephael of the Geological Survey of Israel was crucial to the success of this study. We acknowledge Prof. Abraham Starinsky of the Institute of Earth Sciences of the Hebrew University for field discussions. The research was supported by the Israel Science Foundation grant 683/09 and the Ministry of National Infrastructures, Energy and Water Resources grant 213/17/09.


  1. 1. Garfunkel Z. Internal structure of the Dead Sea Leaky transform (Rift) in relation to plate kinematics. Tectonophysics. 1981; 80: 81–108,
  2. 2. Garfunkel Z. Tectonic setting of Phanerozoic magmatism in Israel: Isr J of Earth Sci. 1989; 38:51–74.
  3. 3. Wald R. Interpretation of the Lower (Southern) Galilee Tectonic Evolution From Oligocene Truncation to Miocene-Pliocene Deformation Using Geological and Geophysical Subsurface Data. University of Haifa, Haifa, Israel. (Ph.D. thesis, In Hebrew, English abstract) 2016; 196 pp.
  4. 4. Freund R, Garfunkel Z, Zak I, Goldberg M, Weissbrod T, Derin B. The shear along the Dead Sea Rift. Philos. Trans. R. Soc. London, Ser. A. 1970; 267:107–130.
  5. 5. Garfunkel Z. The pre-Quaternary geology of Israel. In: Yom-Tov Y., Tchernov E. (Eds.). The Zoogeography of Israel. W. Junk, Dordrecht, the Netherlands. 1988; 7–34 pp.
  6. 6. Stein M, Hofmann A. W. Fossil plume head beneath the Arabian lithosphere?. Earth Planet Sci Lett. 1992; 114(1): 193–209.‏
  7. 7. Shaliv G. Stages in the tectonic and volcanic history of the Neogene basin in the lower Galilee and the Valleys. State of Israel. Ministry of Energy and Infrastructure. Geological Survey of Israel. Rep. GSI/11/91. Ph.D. Thesis, Hebrew University, Jerusalem (in Hebrew, English abstract). 1991; 94 pp.
  8. 8. Kafri U. Neogene to Early Quaternary drainage systems in the Lower Galilee, Israel and their relationship to young tectonics. Isr J of Earth Sci. 2002; 51(2): 79–102.‏
  9. 9. Zilberman E, Sandler A. Coastlines and morphological levels of the Western Lower Galilee–Key of reconstruction of landscape evolution, as response to uplift processes and stable periods. Israel Geological Society Annual Meeting Field Trips Guide, Acco., (in Hebrew). 2013; 18–33.
  10. 10. Segev A. Synchronous magmatic cycles during the fragmentation of Gondwana; radiometric ages from the Levant and other provinces: Tectonophysics. 2000; 325:257–277,
  11. 11. Rozenbaum AG, Sandler A, Zilberman E, Stein M, Jicha BR, Singer BS. 40Ar/39Ar chronostratigraphy of late Miocene-early Pliocene continental aquatic basins in SE Galilee, Israel. Geol Soc Am Bull. 2016; 128(9/10): 1383–1402.
  12. 12. Rosenfeld A, Segev A, Halbersberg E. Ostracode species and paleosalinities of the Pliocene Bira and Gesher Formations (northwestern Jordan Valley): Isr J of Earth Sci. 1981; 30: 113–119.
  13. 13. Raab M, Friedman GD, Spiro B, Starinsky A, Zak I. The geological history of the Messinian (Upper Miocene) evaporites in the Central Jordan Valley (Israel) and how strontium and sulfur isotopes relate to their origin: Carbonates and Evaporites. 1997; 12:296–324,
  14. 14. Schulman N. The geology of the central Jordan Valley. Ph.D. Thesis, Hebrew University, Jerusalem. (in Hebrew, English summary). 1962; 103 pp.
  15. 15. Tchernov E. The biogeographical history of the southern Levant. In: Yom-Tov Y. and Tchernov E. (Eds), The Zoogeography of Israel. Dordrecht: Dr. Junk Publishers. 1988; 159–250.
  16. 16. Rozenbaum AG. The late Miocene-early Pliocene Bira and Gesher Formations: geochronology, depositional environments, and water sources. Ph.D. Thesis, Hebrew University, Jerusalem. (in English. Hebrew abstract). 2017.
  17. 17. Schulman N. The Geology of the Central Jordan Valley. Bull. Res. Counc. Israel. 1959;8G(2–3): 63–90.
  18. 18. Cohen A. The menahemya gypsum in the Jordan Valley: The Messinian Mavqi’im Formation exposed. The Ministry of National Infrastructures. Geological Survey of Israel. Rep. GSI/2/97,;1997. 17 pp.
  19. 19. Rögl F, Steininger FF. Neogene Paratethys, Mediterranean and Indo-pacific seaways; in: Fossils and climate, Brenchey PJ. editors. Geological Journal, special issue. 1984; 11: 171–200.
  20. 20. Rögl F. Mediterranean and Paratethys. Facts and Hypotheses of an Oligocene to Miocene Paleogeography (short overview). Geol Carpath. 1999; 50(4):339–349.
  21. 21. Steininger FF, Wessely G. From the Tethyan Ocean to the Paratethys Sea: Oligocene to Neogene stratigraphy, paleogeography and palaeobiogeography of the circum-Mediterranean region and the Oligocene to Neogene Basin evolution in Austria. Mitteilungen der Österreichischen Geologischen Gesellschaft. 2000; 92: 95–116.
  22. 22. Flecker R., Krijgsman W., Capella W., de Castro Martíns C., Dmitrieva E., Mayser J. P., et al. Evolution of the Late Miocene Mediterranean–Atlantic gateways and their impact on regional and global environmental change. Earth Sci Rev., 2015; 150(2015):365–392.‏
  23. 23. Krijgsman W, Hilgen FJ, Raffi I, Sierro FJ, Wilson DS. Chronology, causes and progression of the Messinian salinity crisis. Nature. 1999;400(6745): 652–655.
  24. 24. Roveri M, Lugli S, Manzi V, Schreiber BC. The messinian sicilian stratigraphy revisited: new insights for the Messinian salinity crisis. Terra Nova. 2008; 20:483–488.
  25. 25. Hsü KJ, Ryan WBF, Cita MB. Late Miocene desiccation of the Mediterranean. Nature. 1973; 242: 240–244.
  26. 26. Koufos G. D. 2003—Late Miocene mammal events and biostratigraphy in the Eastern Mediterranean—In: Reumer J.W.F. & Wessels W. (Eds.)—Distribution and Migration of Tertiary Mammals in Eurasia. A Volume In Honour of Hans De Bruij. -DEINSEA. 2003; 10: 343–371.
  27. 27. Fortelius M, Eronen J, Liu L, Pushkina D, Tesakov A, Vislobokova I, et al. Late Miocene and Pliocene large land mammals and climatic changes. Palaeogeogr Palaeoclimatol Palaeoecol. 2006; 238 (1-4), 219–227.
  28. 28. Blondel J. The nature and origin of the vertebrate fauna. In: Woodward J. (ed.). The physical geography of the Mediterranean Oxford University Press. 2009; 139–163.‏
  29. 29. Zak I. The geology of Mount Sedom. Ph. D. thesis. The Hebrew University of Jerusalem, (in Hebrew, English summary). 1967; 208 pp.
  30. 30. Dicker T. Geology of central Yizre'el Valley*. Isr J of Earth Sci.1969;18:39–69.
  31. 31. Stein M. The evolution of Neogene-Quaternary water-bodies in the Dead Sea Rift Valley. In: Garfunkel Z, Ben-Avraham Z, Kagan E (Eds.). Dead Sea Transform Fault System: Reviews. Modern Approaches in Solid Earth Science. 2014; 6:279–316. ,.
  32. 32. Marcus E, Slager J. The sedimentary-magmatic sequence of Zemah 1 well (Jordan-Dead Sea Rift, Israel) and its emplacement in time and space. Isr. J. Earth Sci. 1985; 34:1–10.
  33. 33. Marcus E.J, Slager S, Ben-Zaken S, Indik IY. Zemah-1 Geological completion Report, Oil Exploration (Investments) Ltd. Tep. 1984; Rep. N° 84/11, 108 pp.
  34. 34. Sneh A, Bartov Y, Rosensaft M. Geological Map of Israel 1: 200,000. Geological Survey of Israel. Sheet 1. 1998.
  35. 35. Braun D. The geology of the Afikim area. Unpublished M.Sc. Thesis, Hebrew University, Jerusalem, (in Hebrew, English abstract). 1992; 93 pp.
  36. 36. Shaked-Gelband D, Stein M, Starinsky A. Lacustrine dolomites from the Lower Galilee, Israel as paleoclimate indicators of the late Miocene East Mediterranean-Levant. 2017. Submitted.
  37. 37. Kidwel SM, Holland SM. Field description of coarse bioclastic fabrics. Palaios. 1991; 6: 426–434.
  38. 38. Zuschin M, Stachowitsch M, Stanton RJ Jr. Patterns and processes of shell fragmentation in modern and ancient marine environments. Earth Sci Rev. 2003; 63: 33–82.
  39. 39. Fürsich FT. The influence of faunal condensation and mixing on the preservation of fossil benthic communities. Lethaia. 1978; 11: 243–250.
  40. 40. Fürsich FT. Palaeoecology and evolution of Mesozoic salinity-controlled benthic macroinvertebrate associations. Lethaia. 1993; 26(4): 327–346.
  41. 41. Clarke KR, Warwick RM. Change in marine community: an approach to statistical analysis and interpretation. Natural Environment Research Council, UK. 1994; 144 pp.
  42. 42. Edelman-Furstenberg Y. Macrobenthic community structure in a high-productivity region: Upper Campanian Mishash Formation (Israel). Palaeogeogr Palaeoclimatol Palaeoecol. 2008; 261: 58–77.
  43. 43. Hudson JD. Aspects of brackish-water facies and faunas from the Jurassic of north-west Scotland. Proceedings of the Geologists' Association 1980; 91(1): 99–105.
  44. 44. Boyer DL, Bottjer DJ, Droser ML. Ecological signature of Lower Triassic shell beds of the Western United States. Palaios. 2004; 19: 372–380.<0372:ESOLTS>2.0.CO;2
  45. 45. Kangas P, Skoog G. Salinity tolerance of Theodoxus fluviatilis (Mollusca, Gastropoda) from freshwater and from different salinity regimes in the Baltic Sea. Estuar. Coast Mar Sci. 1978; 6(4): 409–416.
  46. 46. Amr ZS, Baker MA. Freshwater snails of Jordan. Denisia. 2004; 14: 221–227.
  47. 47. Remane A, Schlieper C. Biology of brackish water. 1971.
  48. 48. Brett CE, Baird GC. Comparative taphonomy: A key to paleoenvironmental interpretation based on fossil preservation. Palaios, 1986; 1: 207–227.
  49. 49. Kidwell SM, Bosence DWJ. Taphonomy and time-averaging of marine shelly faunas. In: Taphonomy, Releasing the Data Locked in the Fossil Record. Allison PA, Briggs DEG, editors. New York, Plenum Press; 1991; 115–209.
  50. 50. Abbott RT. The Gastropod genus Assiminea in the Philippines. Proceedings of the Academy of Natural Sci. of Philadelphia. 1958; 110: 213–278.
  51. 51. Pillay D, Branch GM, Steyn A. Complex effects of the gastropod Assiminea globulus on benthic community structure in a marine-dominated lagoon. J Exp Mar Biol Ecol. 2009; 380(1): 47–52.
  52. 52. Shaked-Gelband D, Edelman-Furstenberg Y, Sandler A, Zilberman E Stein M, Starinsky A. Depositional environments of Bira Formation at Nahal Tavor: Macrofaunal Analysis and Strontium Isotopes. Ministry of Energy and Water Resources. Geological Survey of Israel, Rep. GSI/37/2012, ES-25-2012. 2012; 27 pp.
  53. 53. Heintz E. Les cervidés villafranchiens de France et d’Espagne. Volume I and II Memoires du Museum National d’Histoire Naturelle, n.s. 1970; 22: 1–206.
  54. 54. Malez M, Forsten A. Hipparion from the Bekaa Valley of Lebanon. Geobios. 1989; 22(5): 665–670.
  55. 55. López-Antoñanzas R, Knoll1 F, Maksoud S, Azar D. First Miocene rodent from Lebanon provides the 'missing link' between Asian and African gundis (Rodentia: Ctenodactylidae). Sci Rep. 2015; 5: 12871. pmid:26250050
  56. 56. Horowitz A. The Late Cenozoic stratigraphy and Paleogeography of Israel. Institute of Archeology, Tel Aviv University. 1974.
  57. 57. Horowitz A. The Jordan Rift Valley. A.A. Balkema Publishers, The Netherlands; 2001.
  58. 58. Azanza B, Rössner GE, Ortiz-Jaureguizar E. The early Turolian (late Miocene) Cervidae (Artiodactyla, Mammalia) from the fossil site of Dorn-Dürkheim 1(Germany) and implications on the origin of crown cervids. Palaeobiodiversity and Palaeoenvironments. 2013; 93(2): 217–258.
  59. 59. Ataabadi MM, Fortelius M. Introduction to the special issue "The late Miocene Maragheh mammal fauna; results of recent multidisciplinary research". Palaeobio Palaeoenv. 2016; 96: 339–347.
  60. 60. Geraads D. Large Mammals from the late Miocene of Çorakyerler, Çankiri, Turkey. Acta zoologica bulgarica. 2013; 65(3): 381–390.
  61. 61. Kostopoulos DS. The late Miocene vertebrate locality of Perivolaki, Thessaly, Greece. 9. Cervidae and Bovidae. Palaeontographica Abt A. 2006; 276:151–183.
  62. 62. Bouvrain G, de Bonis L. Ruminants (Mammalia, Artiodactyla: Tragulidae, Cervidae, Bovidae) des gisements du Mioce`ne supe´rieur (Turolien) de Dytiko (Gre`ce). Annales de Pale´ontologie. 2007; 93: 121–147.
  63. 63. Abbazzi L. Cervidae and Moschidae (Mammalia, Artiodactyla) from the Baccinello V-3 assemblage (Late Miocene, Late Turolian, Grosseto, Italy). Rev. Ital. Paleont. Strat. 2001; 107(1): 107–123.
  64. 64. Azanza B, Montoya P. A new deer from the lower Turolian of Spain. J Paleont. 1995; 69: 1163–1175.
  65. 65. Rabinovich R, Gaudzinski-Windheuser S, Kindler L, Goren-Inbar N. The Acheulian site of Gesher Benot Ya‘aqov Vol. III: Mammalian Taphonomy—The Assemblages of Layers V-5 and V-6. Vertebrate Paleobiology and Paleoanthropology series. Spinger, New York. 2012; 1–269 pp.
  66. 66. Amson E, Kolb C. Scaling effect on the mid-diaphysis properties of longbones—the case of the Cervidae (deer). Science Nature. 2016; 103: 58. pmid:27350329
  67. 67. Pitra C, Fickel J, Meijaard E, Groves PC. Evolution and phylogeny of old world deer. Mol Phylogenet Evol. 2004; 33:880–895. pmid:15522810
  68. 68. Croitor R. Description of a new deer species (Cervidae, Mammalia) from the Early Pliocene of Eastern Europe, with a review of early dispersals and palaeobiogeography of the subfamily Cervinae. Neues Jahrbuch für Geologie und Paläontologie - Abhandlungen. 2017; 283 (1): 85–108.