The core area of fruit-tree cultivation: central Jordan Valley (Levant), ca. 7000 BP

Abstract

While it is widely accepted that the five main fruit trees that established horticulture in the late prehistoric period are: olive (Olea europaea), common fig (Ficus carica), grapevine (Vitis vinifera), date palm (Phoenix dactylifera), and pomegranate (Punica granatum), there is much less agreement on where, when, and why this happened. This review paper¹ gathers all recent archaeological and archaeobotanical information on the topic and suggests that all five founders were first assembled into a package in one geographically small region – the Central Jordan Valley. From this core area, knowledge and/or genetic materials were shifted to nearby regions. Yet, it cannot be ruled out that other parallel independent domestications may have occurred in other regions. ¹⁴C dates provided in this study indicate that the beginning of this development is dated at ca. 7000 years cal. BP., earlier than previously considered. It seems that the primary motivation has been related to political and socioeconomic considerations rather than climatological-environmental concerns or other factors. The paper also discusses the cost-effective benefits of simultaneously cultivating several fruit trees. Understanding the early stages of horticulture sheds light on the history of our civilizations, which, according to this study, preceded urbanization or state formation by more than a millennium and a half. A better understanding of the origin and early stages of this development is also of great importance, given the immediate need to adapt horticultural practices to environmental degradation and global climate changes.

Keywords:

• Fruit-tree horticulture
• fruit-tree cultivation
• pollen
• archaeobotany
• Chalcolithic
• olive
• grape
• pomegranate

1. Introduction

The study of the origins of fruit tree domestication is of considerable interest from various viewpoints: from a socio-archaeological perspective, from a fundamental evolutionary understanding viewpoint, as well as from an applied agricultural technology assessment. The main five founders that established fruit trees horticulture are: olive (Olea europaea), grapevine (Vitis vinifera), date palm (Phoenix dactylifera), common fig (Ficus carica), and pomegranate (Punica granatum; Helback 1959; Zohary and Spiegel-Roy 1975; Fall et al. 2002; Zohary et al. 2012; Abbo et al. 2015; Weiss 2015; Fuller and Stevens 2019; Fuller et al. 2023). Today, these fruit trees significantly contribute to human diet in many countries and to global trade.

Agriculture was a major advancement in human history and an irreversible step in most areas around the globe. The shift from hunter-gatherer communities to sedentary farming societies, widely known as the Neolithic Revolution, entailed innovations in plant cultivation and animal husbandry (Zeder 2008). The earliest among these was crop domestication in the Fertile Cresent, primarily cereals and pulses (Childe 1936; Zohary et al. 2012; Asouti and Fuller 2013; Gopher et al. 2021; Zeder 2024). Their composition and seasonality increased the reliability, availability, stability, accessibility, and predictability of food (Zeder 2008, 2011; Abbo et al. 2010). While the earliest evidence of domestication of these crops dates to ca. 11,000/10,500 years ago, the first reliable signs of fruit tree domestication occurred only several millennia later during the Chalcolithic period (Zohary and Spiegel-Roy 1975). It is not clear why people ‘waited’ for about four millennia between the two domestication stages (Fall et al. 2002). Many other aspects regarding the early phases of fruit tree horticulture are also still uncertain or under scholarly debate, including where domestication/cultivation first occurred and whether there were parallel independent domestication events of the same species in different regions.

The terms ‘cultivation’, ‘management’ and ‘domestication’, are often confused (Harris and Fuller 2014; Gopher et al. 2021). In this study, ‘cultivation’ corresponds to the planting of fruit trees, while ‘management’ refers to anthropogenic maintenance of wild fruit trees (i.e. pruning, coppicing). It is assumed that management precedes cultivation and domestication, though this process is not easy to track in the archaeobotanical record (Out et al. 2013; Kabukcu 2018), and it is yet unknown how long it took place. The term domestication is used in this study when there is evidence for an evolutionary change between the wild and the domesticated plant (e.g. genomic evidence, a distinct change in the morphology of the fruit seeds size), or when there is a robust indication of large-scale cultivation of the fruit trees outside their natural habitat. Yet, it should be taken into consideration that domestication is a complex process, and the distinction between ‘wild’ and ‘cultivated’ in terms of practices as well as between ‘wild’ and ‘domesticated’ in terms of genetics and seed morphology is far from being a simple dichotomy (Fuller et al. 2011; Zeder 2011; Gros-Balthazard et al. 2016; Fuller 2018; Gros-Balthazard and Flowers 2021). The understanding of the origin of plant domestication requires the input of multiple disciplines, including botanical archaeology, biogeography, and genetic research. In this study, all relevant information for each of the five founders was gathered in order to address when and where fruit trees were first brought under domestication/cultivation. Currently, the literature indicates that each of the founders had a completely different domestication history (reviewed in this paper), especially in regard to the question of ‘where’. A fresh view into that topic will be provided and it will test a hypothesis that the earliest cultivation of all of the five founders was interlinked and occurred in one very limited core area. I will also try to address the question of ‘why’ – what was the motivation behind this process? I offer new insight into this theme by considering socio-economical factors and evaluating them in light of the available archaeological data. The significance and importance of identifying areas of domestication and of identifying wild ancestors are clear, given the immediate need to adapt horticultural practices to global climate changes and to our degraded environment.

2. Research assumptions and methods

The present review and synthesis build on two basic assumptions: (1) that a fruit tree species’ spatial distribution in Pleistocene–early Holocene geological and in Early Paleolithic–Neolithic archaeological records (Table 1) corresponds to its natural distribution, and (2) that domestication typically begins in a species’ native region (Vavilov and Dorofeev 1992). This paper will gather all relevant available botanical and archaeological information related to the early domestication/cultivation of the five founder fruit trees (an update and elaboration of previous studies: Weiss 2015 and Langgut and Sasi 2023).

Table 1. The chronology of the southern Levant.

Geological period	Archaeological period	Years BP
Late Pleistocene	Epipaleolithic period	23,000–11,700
Early Holocene	Neolithic period	11,700–7500/7200
	Chalcolithic period	7500/7200–6000/5900
Mid-Holocene	Early Bronze Age	5700/5600–4500

Note : Currently, there is a chronological gap between the well-defined periods of the Late Chalcolithic and Early Bronze Age I in the southern Levant (6000/5900–5700/5600 Years BP; e.g. Ashkenazi et al. 2021).

While trying to provide an exact date for fruit-tree domestication and define the regions of early horticulture, it is crucial to acknowledge the affordances and limitations of the various archaeobotanical proxies. Seed and fruit remains in an archaeological site do not necessarily indicate that they were growing nearby; they could have reached the site by transport or commercially. Wood-charcoal remains are more robust indicators of the vegetation in a site’s vicinity. It is widely accepted that the presence of wood-charcoal remains if the site is located outside the species’ natural habitat signifies that this woody plant was most probably domesticated elsewhere and brought there for cultivation (Miller 2008; Willcox et al. 2008; Fuller and Stevens 2019). In such cases, both genetic material and new knowledge (such as horticultural practices) are required. Increasing pollen frequencies can suggest cultivation but not domestication. In addition, the palynological evidence is obligated to provide explanations of whether the increase in pollen ratios is anthropogenic or climate-related. Human activity is indicated when the rise in pollen ratios is not accompanied by an increase of other taxa with the same habitat requirements and when the incline occurs with consistent presence of archaeological and other archaeobotanical findings (e.g. Langgut et al. 2019). Furthermore, while some of the founder trees were proven to be a good marker for cultivation based on their pollen signature (Olea – Mercuri et al. 2013), others are profoundly under-represented because of their pollination mechanism (Vitis – Turner and Brown 2004, Ficus carica – Langgut et al. 2016, and Phoenix dactylifera – Almehdi et al. 2005). Organic residue analysis provides information on micro-remains trapped in or adhered to ancient artifacts. This molecular information is sometimes difficult to interpret (e.g. Fuller and Stevens 2019, p. 270) and can not provide further information on whether the organic residue originated from a wild or domesticated form (Namdar et al. 2015). Lastly, in the case of genetic evidence of fruit species, the distinction between natural and subspontaneous populations is more difficult for several reasons. First, wild and cultivated forms are most often sympatric, favoring gene exchange between them ever since domestication. Second, because fruit trees are relatively long-lived species, few generations separate extant cultivars from their wild ancestors, limiting the amplitude of trait selection, especially as genotypes of domesticated forms were fixed early due to vegetative propagation. Third, since cultivated forms have been spread by human agencies to various regions, naturalized forms previously cultivated and local populations introgressed by introduced cultivated forms could obscure the genetic structure of natural populations (Khadari et al. 2005). Additionally, while DNA data can define areas of potential genetic contributions to the domesticated gene pool, it lacks information on the accurate timing of such events (e.g. Gopher et al. 2021). In this study,¹⁴C dating of young branches/twigs that originated from secure archaeological contexts was therefore used in order to determine when fruit tree cultivation took place. It is based on the assumption that relatively high frequencies of young brunches/twigs will likely derive from seasonal pruning. This common horticultural practice increases fruit yield, among other things. As such, they constitute a short-lived organic material, which is suitable for producing accurate radiocarbon dates (in order to avoid the age of the tree). The criteria for the selection of such a young branch/twig dictated that it would be a relatively small sample whose pith, xylem tissue, and bark were all observed. The ¹⁴C dates were generated by Accelerator Mass Spectrometry (AMS) at the Beta Analytic Laboratory. The radiocarbon age is reported in conventional radiocarbon years (before present = 1950) and calibrated to calendrical years (cal. YBP, by using the program OxCal 4.4, IntCal 20; Bronk Ramsey 2009; Reimer et al. 2020).

This review article will focus only on what is widely accepted by the archaeobotanical community as the five founder species: olive (Olea europaea), common fig (Ficus carica), grapevine (Vitis vinifera), date palm (Phoenix dactylifera), and pomegranate (Punica granatum; Helback 1959; Zohary and Spiegel-Roy 1975; Fall et al. 2002; Zohary et al. 2012; Abbo et al. 2015; Weiss 2015; Fuller and Stevens 2019; Fuller et al. 2023). Other fruit trees such as almond, pistachio and oak will not be discussed here, though there is clear evidence for their exploitation and management during prehistoric times (Asouti and Kabukcu 2014; Asouti et al. 2015; Arranz-Otaegui et al. 2018). These fruit trees should be the focus of a different study, in the same way that recent reviews showed that multiple ‘lost’ crops existed in prehistoric times (Arranz-Otaegui and Roe 2023; Fuks et al. 2024).

3. Results: who, when, and where?

3.1. Olive (Olea europaea)

The earliest archaeological and archaeobotanical evidence for olive oil production originated from the southern Levant and is dated to the Late Pottery Neolithic/Early Chalcolithic period (ca. 7500–7000 BP; Galili et al. 1997, 2018). In the submerged site of Kfar Samir on the Carmel coast (Figure 1a), thousands of crushed olive pits (Figure 2a) were found in association with stone basins and woven baskets (Figure 2b), probably strainers, which presumably were used for olive oil extraction (Galili et al. 1997). In the contemporaneous site of ‘Ain Zippori in the Lower Galilee, olive oil residues were discovered in a pottery vessel (Figure 1a; Namdar et al. 2015), indicating that olive oil was consumed in the northern parts of the south Levantine region at this time. Concerning the processing of table olives, the earliest evidence was found at the submerged site Hishuley Carmel dated to ca. 6600 cal. BP (Figure 1a). It consisted of two small elliptical installations containing thousands of olive pits (Figure 2c and d). Perhaps olive oil production emerged first, and the processing of table olives followed later (Galili et al. 2021).

Figure 1. (a) Phytogeographic map of the southern Levant (Zohary 1973) with sites and archaeobotanical evidence mentioned in the text. M: Mediterranean zone (garrigue, maquis, woodland); IT: Irano-Turanian zone (steppeland); SA: Saharo-Arabian zone (desert); S: Sudanian zone (penetration territory). (b) Map of the southern Levant indicating mean annual precipitation in mm (Srebro and Soffer 2011). Modified after Cavanagh et al. (2022, Figure 1a).

Figure 2. (a) Crushed olive pits from Kfar Samir, interpreted as olive waste (gefet) (photographed by S. Flit). (b) A woven basket found at Kfar Samir, probably an olive oil strainer (akal) (photographed by E. Galili). (c,d) Two stone installations at the submerged site Hishuley Carmel, containing thousands of olive pits that were presumably used for large-scale table-olive processing (photographed by E. Galili).

However, whether these early instances entail the manipulation of wild olive trees or domesticated ones is indeterminate (Galili et al. 1997, 2021; Namdar et al. 2015). All three sites are located in the natural distribution area of wild olives and could have drawn on naturally occurring trees. Conversely, solid indications for purposeful olive tree cultivation were recently identified at Early/Middle Chalcolithic Tel Tsaf in the Central Jordan Valley (ca. 7200–6700 cal. Years BP; Figure 1; Langgut and Garfinkel 2022). They consist of charred olive wood remains found in a region located outside the wild olive’s natural habitat. Olive wood-charcoal remains are absent from the Central Jordan Valley during prehistorical periods (Liphschitz 2007, p. 88–90), and the Tel Tsaf evidence marks their first appearance in the region. Olea europaea is native to the Mediterranean coastal areas (Figure 3), where at least 400 mm of annual precipitation exists (Figure 1b). The area of the Central Jordan Valley is covered by Irano-Turanian stepped vegetation, with an average annual rainfall of 200–300 mm (Figure 1). The additional water supply required for olive horticulture in the vicinity of Tel Tsaf could have been supplemented by irrigation from the nearby Jordan River (Figure 1). Compared with grains and fruits, which can be traded over long distances, wood and charcoal remains of fruit trees represent that the trees were growing in the immediate vicinity of a site (Deckers et al. 2007; Marguerie and Hunot 2007). A few charcoal remains of olive, as well as some olive stones, were also reported in previous studies from Tel Tsaf (Gophna and Kislev 1979; Liphschitz 1988; Graham 2014; Rosenberg et al. 2014).

Figure 3. (a) Olive pollen percentages during the Holocene in the Eastern Mediterranean (note the different percentages of vertical scales). Olea horticulture was indicated when olive pollen ratios increased fairly suddenly, were not accompanied by a rise of other Mediterranean sclerophyllous trees and when the increase occurred with consistent presence of archaeological and archaeobotanical evidence (Langgut et al. 2019). (b) Geographical distribution of wild and cultivated olives throughout the Eastern Mediterranean and suggested dates for the beginning of olive cultivation. Modified after Langgut et al. 2019 (Figures 1, 3 and 8).

A new ¹⁴C radiocarbon date of a young olive branch from Tel Tsaf provided an age of 7000 cal. Years BP (Table 2). Thus, Tel Tsaf encompasses the earliest charred olive wood remains in the Central and Lower Jordan Valley, followed by other Jordan Valley sites dated to the Late Chalcolithic like Abu Hamid, Tell es-Shuna (Neef 1990), Teleilat Ghassul (Zohary and Spiegel-Roy 1975; Meadows 2001), and Pella which also produced large amounts of olive-pressing waste (Dighton et al. 2017). The archaeological and botanical evidence may therefore suggest that olive cultivation began at the Carmel coast and the Galilee, toward the end of the 8^th millennium BP. A few centuries later, at ca. 7000 cal. BP, full-fledged olive cultivation was being practiced at Tel Tsaf, outside Olea europaea’s natural distribution (Figure 3b). Knowledge and genetic material transfer from a nearby area to the Central Jordan Valley must have occurred to accomplish this geographical shift. Numerous olive stones, olive wood remains, and oli presses found at Late Chalcolithic sites in the Golan Heights (Epstein 1993) and Samaria (Eitam 1993) strongly suggest that, by ca. 6000 cal. BP, olive horticulture was well established in these regions and the southern Levant as a whole. The anthracological evidence supports this notion: While during the Neolithic period, the southern Levant charcoal assemblages were characterized by little percentages (or total absence) of olive charcoal remains, since the Early Bronze Age and on, the olive ratios are between 40 and 60% at sites located in the Mediterranean vegetation zone of the region (Liphschitz 2007; Benzaquen et al. 2019; Jin et al. 2024).

Table 2. Radiocarbon dates of charred wood from Chalcolithic Tel Tsaf.

	Lab no.	Dated material	^14C years (BP)	Calibrated age range (68.2% probability) BP	δ^13C ‰	Reference
1	Beta-657404	Young branch of Olea europaea (olive), Area C	6120 ± 30	7160–6950	−24.4	This study
2	Beta-585311	Twig of Ficus carica (common fig), Area C	6060 ± 30	6960–6860	−27.6	Langgut and Garfinkel (2022)

This pattern corresponds well with the results of a comprehensive palynological study spanning the entire Mediterranean Basin (Langgut et al. 2019). It revealed a sudden rise in olive pollen ratios beginning 7000 years ago in the southern Levant (Figure 3a). At the same time, the frequencies of other Mediterranean broadleaved trees (for example, oaks and pistachios) remained more-or-less the same, thus, refuting a climate-related factor. In addition, the rise in pollen ratios was accompanied by archaeological and molecular evidence (Langgut et al. 2019). The earliest of these anthropogenic olive pollen increases was registered at the Sea of Galilee, ca. 7000 cal. BP (Figure 3; Schiebel and Litt 2018), followed by other locations along the Jordan Rift Valley—the Dead Sea (Baruch 1990; Litt et al. 2012), the Hula Valley (Van Zeist et al. 2009), and Birkat Ram (Neumann et al. 2007; Schiebel 2013)—at ca. 7000–6500 cal. BP (Figure 3a). The pollen database presented by Langgut et al. (2019) also indicates that during the early/mid-6^th millennium BP cultivation process occurred in the Aegean (Crete) – whether as an independent large-scale management event or as a result of knowledge and/or genetic material transfer from the southern Levant. The dates provided for anthropogenic increase in olive pollen ratios are: northern Levant – ca. 4800 cal. BP, mainland Italy - ca. 3400 cal. BP, Anatolia - ca. 3200 cal. BP, and the Iberian Peninsula - ca. 2500 cal. BP. In all these regions, the identification of olive cultivation based on the sudden rise in Olea pollen was accompanied by consistent archaeological and archaeobotanical evidence for olive oil production (Langgut et al. 2019). Unfortunately, wild and domesticated Olea pollen is indistinguishable (Figure 4a and b; Supplementary Material 1). This is also true with Olea wood anatomy (Langgut et al. 2019).

Figure 4. Wild and cultivated fossil pollen grains from south Levantine archaeological sites. (a) Wild Olea pollen from the Epipaleolithic site of Jordan River Dureijat. (b) Cultivated Olea pollen identified at the garden of King Herod the Great’s tomb complex at Herodium, first century BC. (c) Wild Vitis pollen extracted from the Epipaleolithic site of Jordan River Dureijat. (d) Cultivated Vitis pollen found at Byzantine Sepphoris. (e) Wild Phoenix dactylifera recovered from ca. 1.6 million years paleolake Zihor, Negev Desert. (f) Cultivated P. dactylifera pollen identified at the peristyle garden of the fortress/palace at Herodium, first century BC. Each bar represents 10 microns (photographed by D. Langgut, I. Fridman, and V. Epstein). Measurements comparing wild and cultivated grains of Olea pollen are presented in the Supplementary Material. Unfortunately, it was not possible to perform such a meticulous analysis on Phoenix dactylifera and Vitis pollen grains.

It is generally accepted that the cultivation of olive trees started with a selection from natural populations of wild Olea europaea L. subsp. europaea VAR. sylvestris (Mill) Lehr (Zohary and Spiegel-Roy 1975; Kaniewski et al. 2012; Barazani et al. 2023). Wild olives reproduce via pollen and spread via seeds (Zohary and Spiegel-Roy 1975). Olive domestication was most probably characterized by the propagation of the ‘better’ phenotypes, such as those with increased yield, bigger fruits, and higher oil content. Though, when exactly vegetative propagation began and whether seedlings preceded it, is yet unknown. In any event, the long history and the extensive distribution of olive culture have resulted in a mixture of wild and feral forms in many Mediterranean habitats (Barazani et al. 2014, 2023). Cross-pollination between wild and cultivated types produced complex populations with various genetic mixtures of domesticated, feral, and wild olive trees (Zohary and Spiegel-Roy 1975; Besnard et al. 2013; Barazani et al. 2023). This situation is further obscured because oleaster plants were, and continue to be, used widely as stock material onto which cultivated clones are grafted (De Candolle 1884; Zohary and Spiegel-Roy 1975; Zinger 1985; Breton et al. 2006; Barazani et al. 2014, 2016). The spread of olive clones by humans in antiquity and their seeds that germinated in various habitats, created additional uncertainty in the cultivar’s identity. In part, this could explain why genetic studies have reached different conclusions about olive domestication’s geographic origin, as well as the number of domestication events (e.g. Besnard et al. 2013 versus Diez et al. 2015). Other explanations for the discrepancy between the genetic investigations might include the use of different methods and sampling issues (Barazani et al. 2023).

A recent review paper about the history of olive provides genetic support to the notion that the southern Levant served as the locus of primary olive domestication (Barazani et al. 2023). Barazani et al. (2014) previously showed that the majority of living old olive trees (90%) in the southeast Mediterranean region belong to the same genetic group associated with the Souri cultivar. The latter is typical to the southern Levant and occupies most of the traditional rain-fed olive orchards (Zinger 1985; Ben-Ari et al. 2014). It is adaptable to different climatic habitats, including semi-arid environments (200–400 mm of average annual precipitation) and shallow and stony marginal soils (Ben-Ari et al. 2014). Recent studies also pointed to the higher drought tolerance of the Souri cultivar (Tugendhaft et al. 2016; Barzilai et al. 2021). That means that the selection of the Souri cultivar enabled the expansion of the olive cultivation zone in the southeast Mediterranean into more arid habitats (e.g. Tepper et al. 2022). Indeed, this study provides the earliest ¹⁴C date of olive cultivation outside its natural habitat (Table 2) at the site of Tel Tsaf, which is located in a semi-arid steppe environment, receiving today 200–300 mm of average annual rainfall (Figure 1b). Baruch (1990), based on pollen records recovered from the Jordan Valley, suggested that Olea cultivation between the Chalcolithic and the Roman/Byzantine eras was concentrated at lower elevations. This proposal is based on the observation that the pollen ratios of deciduous oak remained high during this time interval and decreased only after the classical periods (Baruch 1990). This suggestion is also in accordance with new optically stimulated luminescence dates for the beginning of terraced landscapes in the Mediterranean parts of the southern Levant (Gadot et al. 2018).

3.2. Grapevine (Vitis vinifera)

It is suggested in the literature that the domestication process of Vitis started in southwest Asia and the varieties obtained were successively spread and cultivated in different areas (Zohary and Spiegel-Roy 1975; Arroyo-García et al. 2006; Grassi and De Lorenzis 2021; Bouby et al. 2023; Dong et al. 2023). However, whether it occurred once or whether successive domestication events occurred independently is highly debated. Introgression events, breeding, and intense trade across the Mediterranean Basin over the last several millennia produced ambiguous genetic relationships (for recent reviews, see Grassi and De Lorenzis 2021 and Meiri and Bar-Oz 2024). Today, several thousand varieties can be distinguished, and they are generally classified into two main categories: table (fruits consumed fresh or dried) and wine grapes. It was suggested that the varieties have been generated since grapevine domestication by vegetative propagation and by crosses (e.g. Myles et al. 2011), yet seedlings cannot be ruled out, especially during the early steps of the process.

The archaeobotanical evidence is also inconclusive regarding where in southwest Asia and when table and wine grapes were brought under domestication since there is a considerable morphological overlap of domesticated and wild seeds and berries. In addition, wild species grow near sites with early evidence of grapes (Miller 2008; Weiss 2015). Although several recent studies have indicated morphological criteria for distinguishing wild and domesticated grape pips (Terral et al. 2010; Orrù et al. 2013; Pagnoux et al. 2015, 2021; Valamoti 2015; Bonhomme et al. 2021, 2022; Bouby et al. 2021; Chen et al. 2022; Valera et al. 2023), these differences emerged over centuries and millennia and, therefore, cannot provide an accurate date for domestication. Besides grape pips, other botanical remains such as grapeskins (Fuks et al. 2020), wood (Benzaquen et al. 2019) and pollen (Turner and Brown 2004) are usually underrepresented in archaeological contexts. Hence, the lack of these types of botanical remains in the archaeological record should not be taken as evidence of absence, but the presence of any one of these remains is highly suggestive that Vitis was growing nearby or at least consumed/processed (Miller 2008; White and Miller 2018; Fuks et al. 2020). For instance, Vitis is usually underrepresented in wood-charcoal assemblages since it possesses a low density, and therefore, the weak constitution of lianas deteriorates easily (0.40 g/cu cm; Crivellaro and Schweingruber 2013, p. 569). Also, due to its low density, grapevine wood is considered a poor-quality fuel material and is unsuitable for use in construction and the preparation of wooden artifacts (Figure 5); hence grapevine remains are relatively rare in wood-charcoal assemblages (Deckers et al. 2023). Accordingly, when it does occur, even in low frequencies, it indicates that Vitis was growing nearby (Miller 2008; Bouby et al. 2023). Likewise, grape pollen is under-represented in palynological spectra. Most domesticated grapes are monoecious (producing both male and female flowers) and are self-pollinated (flowers contain both pistils and anthers), resulting in low pollen dispersal efficiency. Experiments demonstrated that grape pollen abundance exponentially declines with distance from vineyards and that a relative abundance of 2% grape pollen in an assemblage is strong evidence for nearby grape plants (Turner and Brown 2004). Despite the changes in pollen dispersal mechanism after domestication, wild and domesticated pollen grains are indistinguishable (Figure 4c and d, but see also Mercuri et al. 2021), and this is also the case with the wood anatomy. Yet, many grape varieties are characterized after their domestication by an increase in berry and bunch sizes, a rise in sweetness and acid content, a variation in berry color and shape and more regular yields and perfect flowers (This et al. 2006). Bouby et al. (2023) use the presence of pedicels as evidence for winemaking. What is clear is that the period when grapevines spread beyond their natural wild distribution, can be taken as indicative of cultivation, if not domestication (Miller 2008; Fuller and Stevens 2019).

Figure 5. A grapevine plantation during winter in the Judean Mountains near Jerusalem. Due to grapevine wood’s low density, wood-charcoal remains are rarely found in dendroarchaeological assemblages. Grapevine wood is considered a poor-quality fuel and unsuitable for construction or preparation of wooden artifacts (photographed by T. Langgut).

Many researchers consider the area of the southern Caucasus, between the Caspian and Black Seas, as the most likely origin of grape cultivation (Olmo 1995; McGovern et al. 1996; Myles et al. 2011; Riaz et al. 2018). The region is notable for the vast diversity of wild populations and cultivars (Bouby et al. 2021) and provides the earliest evidence of wine production, consisting of chemical residues in early 8^th-millennium BP pottery vessels (McGovern et al. 1996, 2017; but see critique by Fuller and Stevens 2019, p. 270). However, as in the case of the early production of olive oil (Galili et al. 1997; Namdar et al. 2015), it is unclear whether the wine was produced from wild or domesticated grapes (Miller 2008; Bouby et al. 2021). Moreover, evidence of grapes (wood, pollen, and fruit remains) did not enter the area’s archaeological record until 3000 years later, in the mid-5^th millennium BP (Miller 2008). At the late-7^th millennium BP Dikili Tash site in northern Greece, grape juice and fermentation markers were detected inside a jar that was also associated with archaeobotanical remains of grape pressings (Valamoti 2015; Garnier and Valamoti 2016). Tartaric acid, which can serve as an indicator of wine production, was also recorded at Areni-1 in Armenia, dated to ca. 6000 BP (Barnard et al. 2011). The morphology of the grape pips recovered from the site indicates that they are an intermediary form between wild and domestic, suggesting that they may represent the initial stages of Vitis domestication (Areshian et al. 2012). A chemical residue pointing to the presence of wine was also extracted from a jar at Godin Tepe, dated to the 6^th millennium BP. The site is located in western Iran, some 400 km from the present wild Vitis territory (Badler et al. 1990). In the Northern Levant, the find of underdeveloped grape pips at Kurban Höyük dating to the end of the 6^th millennium BP suggests Vitis cultivation (White and Miller 2018), and matches with the morphometric analysis on seeds which identified grape cultivation by the latest in the 5^th millennium BP in the area of the Middle Euphrates (Valera et al. 2023).

In the southern Levant, the wild grapevine occurs in wet habitats such as the upper Jordan River and the Hula Valley but is not a very common component of the natural vegetation. The oldest grape remains from the Jordan Valley originated from the 780,000 years old Gesher Bnot Ya’ackov site, where pollen (Van Zeist and Bottema 2009), wood remains (Goren-Inbar et al. 2002) and seeds (Melamed et al. 2016) were identified. A few pips of Vitis were recovered from the 23,000 years old fishercamp of Ohalo II on the south-eastern shore of the Sea of Galilee (the area is considered to be the beginning of the Central Jordan Valley; Figure 1; Kislev et al. 1992) and several fossil pollen grains (Figure 4c) and wood remains were recovered from the Epipaleolithic site on the southern shore of paleolake Hula, Jordan River Durijat (Figure 1a; Langgut et al. 2021). Fossil pollen grains were also extracted from Early Holocene Sea of Galilee palynological record (Schiebel 2013). The wild grapevine is much more widespread in wetter habitats, such as in the river valley of the Zagros and Pontus ridges in Turkey, in comparison to the Levant wet environments. Indeed, most wild grapevine habitats have annual precipitation levels exceeding 600 mm, and the southern Levant is on the cusp and below this figure (Figure 1b). Nevertheless, ecological studies have shown that the wild populations are remarkably adaptable, capable of growing in a broad range of habitats and various soils, including along seasonal rivers in closed forests, forested wetlands, and sand dune shrublands (Zohary 1973; Naqinezhad et al. 2018). The minimum threshold of annual rain becomes significantly less important when constant irrigation is provided (Chen et al. 2022). Domesticated grapevines capitalized on the adaptability to various habitats and/or the availability of irrigation. They were carried over into drier regions—as far as the hyper-arid Negev Desert—and supported the southern Levant’s establishment as a center of ancient winemaking that lasted until the collapse of the Byzantine Empire (Fuks et al. 2020; Cohen et al. 2023). When this began in the southern Levant region is unclear. But, the discovery of pips, charred berries, and wood in Early Bronze I Jericho (Table 1; Hopf 1983; Western 1983), a site too dry and too warm to support wild Vitis (Figure 1b), offers a terminus ante quem for this event. The presence of fruit remains as well as wood from the grape plant in a site where wild grapevine could not grow, provides solid proof for horticulture-type domestication (Miller 2008; Weiss 2015). Other Early Bronze Age sites from the region that include grapevine remains, are (Figure 1a): Bet Yerah (Berger 2013), Numeira and Bab edh-Dhra (McCreery 1979), Arad (Hopf 1978), and Lachish (Helbaek 1958). Grape remains also appear in Egypt at this time (Weiss 2015, and references therein).

The northern part of the Jordan Valley may have served as an independent center of grapevine domestication, though more data are required to strengthen this suggestion. It is proposed in this study that during the Chalcolithic period, domestication spread southwards, reaching Jericho by the mid-6^th millennium BP. To accomplish this geographical shift, a transfer of both knowledge and genetic material from the Northern Jordan Valley to the central and southern segments of the valley must have occurred. This hypothesis is supported by Sivan et al.’s (2020) observation that the Levantine grapevine varieties have a distinct genomic background. However, they also stress that it remains unclear whether ancient Levantine varieties stemmed from extensive gene flows across the Near East or directly domesticated Levantine wild grapevines. A recent genetic study of whole genomic sequencing of ca. 4000 accessions of wild and domestic grapevine may clarify the geographic origin of Vitis domestication. The results point to two domestication events: the Caucasus wine varieties as local, restricted, domestication, and the wild south Levantine population domesticated as table grapes and became the origin of most old-world table domestic varieties (Dong et al. 2023). Later, introgression between the Levantine grapevine and wild Turkish populations led to wine varieties, which spread around and went through additional introgressions with European wild varieties (Dong et al. 2023). As in the case of olives, changes took place in the domesticated grapevine gene pool by hybridization with local wild populations as it shifted from the Near East into Europe (Myles et al. 2011; Weiss 2015; Bouby et al. 2023; Cohen et al. 2023; Meiri and Bar-Oz 2024). A study conducted by Valera et al. (2023) analyzed the morphometric shape of grape seeds from the northern Levant (Middle Euphrates region). The study revealed that during the 5^th millennium BP, there were several grape varieties, including a local variety of domestic grape which was likely a hybrid of Asian and South Caucasian domestic vines (Valera et al. 2023). The degree to which local wild sylvestris from Central and Western Europe contributed genetically to Western European vinifera cultivars is still under scholarly debate (Myles et al. 2011; Grassi and De Lorenzis 2021). Secondary domestication events may have occurred anywhere in the area of the natural distribution of the wild grapevine, ranging from western Europe to the Himalayas and around most of the Mediterranean Basin. Parallel domestication events may also have occurred in this area without any role of introduced cultivars, but further datasets are required to corroborate this topic (Pagnoux et al. 2021). In light of the review above, this study suggests that the northern Jordan Valley served as one of the areas where independent grapevine domestication occurred.

3.3. Date palm (Phoenix dactylifera)

The origins of domesticated date palms are still controversial. Researchers have developed different theories about the wild progenitor species and debated the location of center(s) of origin, the number of domestication events, and how hybridization may have contributed to the diversification of the date palm (Gros-Balthazard and Flowers 2021, for a recent review). Phoenix dactylifera is native and economically important in arid and semi-arid regions of the southern Mediterranean Basin, north Africa, the Sahara and southwest Asia (Zohary et al. 2012, p. 132), yet there are some questions regarding its original natural distribution. Present-day populations of date palm consist of wild forms, segregated escapees, and domesticated clones, which are all genetically interconnected by occasional hybridization (Figure 6a; Zohary et al. 2012, p. 132; Gros-Balthazard and Flowers 2021). Since it is also difficult to distinguish wild from domesticated date palms in the archaeobotanical record based on the seed morphology (e.g. Weiss 2015), scholars have been relying on a marked increase in the ratios of date kernels to detect domestication. In a recent study on that topic, Gros-Balthazard et al. (2016) found that seed size is uninformative in differentiating feral from wild date palms at the intra-specific level, as both may display small seeds because of constrained environmental conditions. Yet, seeds from domesticated individuals develop larger seeds as a consequence of selection and cultivation practices. These differences, however, developed over centuries and millennia and cannot be used to determine the exact date of domestication. Though based on pollen morphology, wild and domesticated P. dactylifera are indistinguishable (Figure 4e and f), a profound increase in date palm pollen can be used as a marker for cultivation in the same way it is used for other fruit trees (Bottema and Woldring 1990; Mercuri et al. 2013). The palynological proxy is specifically useful since date palm trees are characterized by low pollen dispersal efficiency (Almehdi et al. 2005), and therefore a significant increase in P. dactylifera pollen ratio in locations where both Pleistocene and Holocene palynological records are available can be used to trace anthropogenic activity.

Figure 6. (a) Geographical distribution of wild and feral forms of date palm, Phoenix dactylifera, modified after Zohary et al. (2012, map 18). In South Arabia, Sudan, Eritrea, Somalia, and Senegal, they sometimes cross with the Native P. reclinata; in the Indus Basin, they occasionally cross with P. sylvestris. (b) A young date palm plantation on the northwestern coast of the Dead Sea (photographed by T. Langgut).

The Quaternary sedimentological record indicates that date palms thrived in the hot and dry parts of southwest Asia and the southern Mediterranean long before the beginning of plant domestication. The earliest remains originated from 1.6 million years ago, paleolake Zihor, located today in the arid region of the southern Levant (Figure 4e; Fridman 2023). Pleistocene presence of date remains was also detected at Jebel Faya (United Arab Emirates; Bretzke et al. 2013, though an identification to the species level was not possible), in the southeast Kingdom of Saudi Arabia (Groucutt et al. 2015) and at the Egyptian oasis of Kharga (Gardner 1935). The occurrence of date palm remains at prehistoric sites may point to the long history of human exploitation of this plant: Late Middle Paleolithic sites of Shanidar cave in Iraq (Solecki and Leroi‐Gourhan 1961; Miller-Rosen 1995; Madella et al. 2002; Henry A et al. 2011) and Tor Faraj rock shelter (southern Jordan; Henry et al. 2004) as well as Epipaleolithic Ohalo II (Liphschitz and Nadel 1997). Date palm remains dated to >12,000 years BP were also reported from southern Iraq (Altaweel et al. 2019) and from Pre-Pottery Neolithic south Levantine sites - off the Levantine coast at Atlit-Yam (Liphschitz 2007, p. 39), in the Jordanian Desert at Ghuwayr 1 (Simmons and Najjar 2003) and Tell Wadi Feinan (Jenkins et al. 2011). 7000-year-old evidence of date palm exploitation by humans has also been documented at two sites on the Gulf coast: Sabiyah, Kuwait (Parker 2010) and Dalma Island, United Arab Emirates (Beech and Shepherd 2001). Despite the fact that all these types of evidence are sporadic and inconsistent, it seems likely that wild P. dactylifera is native to oasis areas in Western Asia, although the precise distribution is unknown. A clear rise in the frequencies of date kernels was reported only since the Chalcolithic period. Such evidence was registered at South Levantine sites from the Dead Sea region, such as Tuleilat Ghassul (Zohary and Spiegel-Roy 1975) and the Cave of the Treasure (Zaitschek 1961). An abundance of presumably domesticated date kernels was also recorded in the ca. 6000-year-old Ubaidian horizon at Eridu, Lower Mesopotamia (Gillett 1981). The occurrence of these remains close to current and past habitats of wild date populations emphasizes the need to establish their domestication status beyond the sheer quantity of kernels (Zohary et al. 2012, p. 134). Moreover, dates were widely traded and consumed. Consequently, the seeds were dispersed through long distances, obscuring the species’ original distribution (Figure 6a; Barrow 1998; Tengberg 2012; Gros-Balthazard and Flowers 2021). We thus have only little data about the date palm’s origins, domestication, historical biogeography, and evolutionary history (Abbo et al. 2015, p. 338; Gros-Balthazard et al. 2017; Gros-Balthazard and Flowers 2021). As a result, efforts to indicate sites and dates of domestication tend to produce loose and indefinite conclusions (Méry and Tengberg 2009; Weiss 2015; Zehdi-Azouzi et al. 2015; Gros-Balthazard and Flowers 2021).

The cultivated tree can be propagated from seeds and, unlike its wild relative, can be vegetatively propagated from offshoots (suckers) at the base of the plant (Figure 7; Janick 2005, p. 276), hence preserving fruit quality across generations (Gros-Balthazard and Flowers 2021). Notably, all species of Phoenix are dioecious (Barrow 1998), which called for early acknowledgment of sex, already during the days of Hammurabi (1792–1750 BC; Zohary et al. 2012, p. 131). Assyrian murals dating to the tenth century BC, are illustrated by representations of artificial pollination of date palms (Paley 1976; Bryant 1990). They also point to the ancient understanding of the value of pollination and its advanced culture in ancient Mesopotamia.

Figure 7. An offshoot (a sucker) of a juvenile date palm’s axillary bud groomed for vegetative propagation at a date palm plantation in the Central Jordan Valley (photographed by D. Langgut). When at ground level or when surrounded by boxes (in the past), or wrapped in plastic sheets (today), the offshoots are enclosed in soil or potting mixture, which allows them to develop an adventitious root system; after three to five years, it can be removed from the mother tree and replanted.

Although it is difficult to identify wild stands of date palm, Zohary et al. (2012, p. 132) proposed that at the southern parts of the Dead Sea Basin—and, at the southern base of the Zagros Range facing the Persian Gulf, wild-type dactylifera palms are still present. During the Pleistocene, the Dead Sea palynological sequence indicates a sporadic and inconsistent appearance of date palm pollen (Chen and Litt 2018, Appendix A). A more significant presence was documented in the Dead Sea pollen record since ca. 6500 BP (Litt et al. 2012, Figure 3). A profound increase in the frequencies of date kernels was observed at the same time in Chalcolithic southern Levant sites, near the Dead Sea (Zaitschek 1961; Zohary and Spiegel-Roy 1975). Taking into consideration parallel independent domestication events, it is suggested in this study that the area of the Dead Sea and Jericho (Figure 1), located at the southernmost section of the Central Jordan Valley, may have served as one of the regions of P. dactylifera domestication. Interestingly, in the classical periods, the dates that were grown in the Dead Sea region were famous worldwide for their qualities. Theophrastus (fourth century BC) described the area around the Dead Sea as being famed in antiquity for the variety of dates grown in the orchards there (Theophrastus 1916, Historia Plantarum II.6:6-7). Pliny the Elder (1952; first century AD) discusses the Judaean palm (Natural History XIII.6:16)—specifically the dates coming from Jericho—calling them the most well-known (Natural History XIII.9:44). During modern times, the date palm is considered one of the most profitable cash crops in the Dead Sea area (Figure 6b; Abu-Qaoud 2015; Cohen and Glasner 2015).

3.4. Pomegranate (Punica granatum)

The wild pomegranate types occupy a wide range of elevations from below sea level to ca. 2000 m above sea level, indicating that while genotypes tolerate high temperatures, others withstand low ones (Browicz 1996). Wild forms of P. granatum grow in masses in the south Caspian belt, northeastern Turkey, and Albania and Montenegro (Zohary et al. 2012, p. 134), and it is assumed that the tree was cultivated there (Abbo et al. 2015, p. 340). The southern Levant was ruled out as a possible domestication site because the tree was thought not to occur there naturally (Zohary et al. 2012, p. 135). However, this study suggests otherwise. Pomegranate pollen recovered from 780,000-year-old Gesher Benot Ya‘aqov speaks for the species’ long history in the region (Van Zeist and Bottema 2009). Significantly, it is insect-pollinated and characterized by very little wind dispersal ability due to its heaviness, rendering a long distant origin highly unlikely (Keogh et al. 2010). Recently, several waterlogged P. granatum wood remains were discovered in nearby Epipaleolithic (ca. 23,000–11,000 cal. BP) Jordan River Durijat (Figure 8), indicating the presence of wild pomegranate in the Northern Jordan Valley. At Pre Pottery Neolithic B Nahal Oren, a carbonized pomegranate seed was found (Noy et al. 1973). During the Chalcolithic period, P. granatum fruit remains were discovered in two Judean Desert caves (Cave V/49 and Cave of the Treasure; Melamed 2002 and Zaitschek 1980, respectively). By the Early Bronze Age (Table 1), pomegranate remains were more prevalent in the region, probably an indication of its cultivation: Early Bronze Age Jericho (Hopf 1983; Western 1983), Arad (Hopf 1978), and Bet Yerah (Mor 2022). Based on these regional archaeobotanical data, it seems that pomegranate was native to the flora of the Jordan Valley, though with a very limited distribution.

Figure 8. Scanning Electron Microscope (SEM) images of charred-wood sections of Punica granatum recovered from the Epipaleolithic site Jordan River Dureijat (Northern Jordan Valley). (a) P. granatum transverse section (vessels diffuse; in radial multiples of 2–4; 30–40 μm average diameter). (b) P. granatum tangential longitudinal section (vessels ca. 20–40 μm in diameter; Uniseriate rays can be seen). (c) P. granatum tangential longitudinal section (vessel ca. 40 μm in diameter; vessel pits ca. 3–5 μm in diameter). Photographed by M. Cavanagh with a Tescan VEGA3 LMH SEM.

The pomegranate is a minor crop in traditional Mediterranean horticulture (Zohary et al. 2012, p. 134), constituting a regular component in mixed orchards and home gardens (bustan). However, its fruits are not amenable to simple preservation, which is probably why it became less economically significant in comparison to the other founders (Bonfil and Hadas 2011; Abbo et al. 2015, p. 365).

3.5. Common fig (Ficus carica)

It is unclear and sometimes contradictory where and when figs were domesticated. The wild type of fig belongs to a group of genetically and reproductively compatible forms, the outcome of the spread of feral types of the domesticated common fig among its wild types and innumerable crosses between its domesticated, feral, and wild forms (Zohary 1973; Lev-Yadun 2022). Relatively vast natural populations of Ficus carica are found in the Colchic district of north-eastern Turkey and the Hyrcanian district of northern Iran. However, populations of what is considered and sometimes proved to be wild common figs are found in many places all over the Mediterranean Basin, as attested by archaeobotanical remains (Lev-Yadun 2022) and the genetic evidence (Khadari et al. 2005). Moreover, the existence of wild common fig populations in various sites around the Mediterranean Basin provided the pollinating wasp (Blastophaga psenes; Galil and Neeman 1977) and valuable common fig genetic resources when common fig horticulture spread (Khadari et al. 2005; Lev-Yadun 2022). Unfortunately, the ancient distribution and routes of the spread of domesticated types remain unknown. The genetic study conducted by Khadari et al. (2005) shows that fig populations are structured into three clusters: Balearic, East, and West Mediterranean gene pools. Balearic populations’ low diversity and high differentiation indicate an ancient origin and the presence of natural populations in this region before domestication. The significant genetic differentiation between the East Mediterranean and the West Mediterranean may also be attributed to the diversification of common fig across the Mediterranean basin preceding domestication. As opposed to this, Italian island populations appear to be the result of introduced cultivated figs, since they present continental haplotypes (Khadari et al. 2005).

Within the archaeobotanical record, it is impossible to distinguish wild from domesticated figs based on seed morphology and wood anatomy (Zohary et al. 2012, p. 129; Weiss 2015). Pollen is under-represented in archaeological and geological records since F. carica is pollinated by an elaborate symbiosis with a particular species of a wasp; the flowers are practically invisible as they bloom inside the syconium (Galil and Neeman 1977). Thus, virtually no fig pollen is released into the atmosphere, and hence, the fossil fig grains are found only in rare and special archaeological contexts (e.g. Langgut et al. 2016).

Some of the earliest evidence of common fig originated from prehistoric archaeological sites across the Jordan Valley (Figure 1), where natural stands can still be found today. The 780,000-year-old Gesher Benot Yaʻaqov site produced fig seeds (Melamed et al. 2011, 2016) and wood specimens (Goren-Inbar et al. 2002). Fig remains were also found in Epipaleolithic (Weiss 2017) and Neolithic (Kislev 1997; Kislev et al. 2006) sites along the Jordan Valley, as well as in other Early Neolithic Levantine sites (Van Zeist and Bakker-Heers 1982; Rollefson et al. 1985; Kislev and Hartmann 2012; Hartmann-Shenkman et al. 2015), and throughout southwest Asia (Miller 1991, table 2; Weiss 2015; and references therein). All these data indicate that during prehistoric times the fig was a natural element in the Jordan Valley and widely distributed over many other parts of the Mediterranean basin.

Recently, Langgut and Garfinkel (2022) described an assemblage of common fig charcoal remains from Tel Tsaf, dated to 7000 cal. BP (Table 2), and suggested that it constitutes one of the earliest recorded instances of common fig management/cultivation (Langgut and Garfinkel 2022). This proposal is based on the identification that the majority of the fig charred wood remains originated from young branches and twigs that may have derived from pruning. Like many deciduous fruit trees, figs require yearly pruning before terminating their winter dormancy (Flaishman et al. 2008). These seasonal pruning activities were (and still are) a standard practice in fruit tree horticulture: pruning controls the tree’s shape, allows sunlight to reach all branches, restrains its aggressive growth (and size), and facilitates efficient fruit harvest. This provides common fig growers with ample plant material, enabling propagation from cuttings and fuel material (Flaishman et al. 2008; Lev-Yadun 2022). The trimmed branches are removed to prevent the spreading of fungi and pests onto healthy trees, serving as a readily available fuel source at sites adjacent to the orchards (Jin et al. 2024). This practice is still common in traditional societies (Hobbs 1989, p. 53; Andersen et al. 2014). The presence of fig seeds at Tel Tsaf supports the evidence that originated from the charred wood remains (Gophna and Kislev 1979, p. 113). Though it is unclear whether the fig twigs represent pruning practices of wild trees or cultivated ones, it is worth noting that fig wood rarely occurs in the archaeobotanical record. This is probably due to its limited usefulness: Fig does not provide long and sturdy beams and is unknown to have been traded for other purposes. Therefore, when fig wood remains are found, it may be inferred that fig trees grew in the site’s vicinity (Lev-Yadun 2022). A substantial number of young fig branches were also identified in Tel Bet Yerah’s charcoal assemblage (located about 30 km north of Tel Tsaf; Figure 1a), suggesting the continuity of fig management/cultivation in the region since the Chalcolithic period to the Early Bronze Age (Mor 2022). Archaeobotanical evidence of olive, grape, and pomegranate was also discovered at Tel Bet Yerah (Berger 2013; Mor 2022). Wood remains of fig were also found at Early Bronze Jericho, alongside wood remains of pomegranate, date palm and grape (Western 1983), suggesting that by the Early Bronze Age, the fruit tree package was already well established in the Central Jordan Valley.

Several hypotheses have been proposed regarding the timing of the domestication of the common fig. The most widely accepted theory proposes that F. carica was domesticated, with other founder fruit trees, during the Chalcolithic period, some 6000 years ago (Weiss 2015). Another theory proposed by Kislev et al. (2006) claimed that the common fig was domesticated in the Lower Jordan Valley during the Pre-Pottery Neolithic A, slightly before the beginning of grain crop agriculture. Nevertheless, this hypothesis was rejected since the parthenocarpic female figs discussed by Kislev et al. (2006) can also occur naturally (Lev-Yadun et al. 2006; Denham 2007; Zohary et al. 2012; Abbo et al. 2015; Weiss 2015). There is also disagreement over whether F. carica domestication occurred at a specific time and place, gradually spreading throughout the Mediterranean Basin, or whether it resulted from multiple unrelated cultural events. This indetermination results from the female F. carica’s clonality and the lack of anatomical differences between wild and domesticated types, making distinguishing between primary and secondary locations difficult to detect (Lev-Yadun 2022). Like the other four founders of the fruit tree package, fig trees can be easily propagated from branch or stem cuttings (Zohary and Spiegel-Roy 1975; Flaishman et al. 2008). Branch cuttings with several nodes are removed from the parent tree in late winter and planted in wet soil, in which it will readily produce roots and resume growth when temperatures rise in early spring after bud break. Such trees, planted from branch cuttings, can produce commercial fruit within two to three years, depending on soil fertility and other environmental factors (Stover et al. 2007; Flaishman et al. 2008; Lev-Yadun 2022). Even so, seedlings cannot be ruled out during the early phases of domestication, as with the other founders.

4. Discussion

4.1. The core theory: assembling the fruit tree package

This study proposes that the five founders of fruit tree horticulture were assembled into ‘one package’ in a restricted geographical area: the Central Jordan Valley. A parallel model of one core area for the transition to agriculture was raised by Lev-Yadun et al. (2000). The Neolithic model depicts a core area in the northern Levant for all seven founders of the Neolithic package that were domesticated during one short, single, knowledge-based event (for a different view: Willcox 2005; Fuller et al. 2011; Riehl et al. 2013).

The suggestion in this study that the Central Jordan Valley served as the core area for the beginning of fruit tree horticulture is based on the geographical overlap of the first cultivated fruit trees’ distributions. The geological, archaeobotanical and climatological evidence (Figures 1 and 3b) indicates that olive is the only species among the five founders that is not native to what is suggested here as the core area. However, the earliest worldwide evidence for olive cultivation derives from 7000-year-old Tel Tsaf (Table 2; Langgut and Garfinkel 2022). Since olive does not occur naturally in the Central Jordan Valley, the exchange of both knowledge and genetic material was required, most probably from the Carmel Coast, where evidence for Late Pleistocene-Early Holocene Olea natural distribution exists (e.g. Kadosh et al. 2004), as well as the documentation of the earliest production worldwide of olive oil already at ca. 7500 cal. BP (Galili et al. 1997, 2018). Interestingly, evidence for direct contacts between the Central Jordan Valley and the Mediterranean Coast at this time is provided by Dead Sea bitumen on sickle blades from Atlit-Yam (Oron et al. 2015) and Mediterranean Sea shells at Tel Tsaf (Rosenberg et al. 2023).

Although the contemporary southern Levant features natural stands of date palm only in the Dead Sea area (Zohary et al. 2012, p. 132), its distribution spanned all parts of the Jordan Valley (north, center, and south) during the Pleistocene (Horowitz 1986; Liphschitz and Nadel 1997; Jenkins et al. 2011; Schiebel 2013; Chen and Litt 2018; Langgut et al. 2021; Fridman 2023). Remarkably, while the Holocene Dead Sea pollen diagram recorded that date palm cultivation had begun at ca. 6500 cal. BP (Litt et al. 2012, Figure 3), nothing of the sort is observable in the Sea of Galilee diagram (Schiebel and Litt 2018). It therefore appears that since the early Holocene date palm distribution was more restricted to the oasis areas near the Dead Sea. Interestingly, the same palynological record indicates that the advent of olive cultivation in the Dead Sea area coincided with date palm cultivation.

Pleistocene and early Holocene grapevine archaeobotanical remains were recovered from the Northern (Van Zeist and Bottema 2009; Goren-Inbar et al. 2002; Melamed et al. 2016; Langgut et al. 2021) and Central Jordan Valley (Kislev et al. 1992; Liphschitz and Nadel 1997; Schiebel 2013). At the same time, and as detailed above, date palm remains were observed in the Central Jordan Valley (Liphschitz and Nadel 1997; Schiebel 2013), as well as in the Southern Jordan Valley (Jenkins et al. 2011; Chen and Litt 2018; Fridman 2023). Given these species’ distinct northern and southern distributions, their concurrence here constitutes the Central Jordan Valley as their only point of convergence.

Naturally occurring pomegranate was recorded only in the Northern Jordan Valley during prehistorical times (Figure 8; Van Zeist and Bottema 2009), but by the Early Bronze Age seems to have been cultivated in the Central (Mor 2022) and southern Jordan Valley (Western 1983). Similarly, the common fig is also native to the Northern Jordan Valley (Goren-Inbar et al. 2002; Melamed et al. 2011; Langgut et al. 2021) but has been cultivated in the Central Jordan Valley since 7000 BP (Tables 2 and S2; Langgut and Garfinkel 2022), and on, as indicated by the wood-charcoal remains recovered from Bet Yerah (Mor 2022).

While olive, common fig, pomegranate, and grapevine are native to Levantine Mediterranean climatic zones, the date palm is best adjusted to arid conditions. To yield high-quality fruits, the palm tree requires high temperatures and low humidity; trees that grow in areas with mild summers, like the Levantine Mediterranean coast, are likely to bear low-grade fruits. Thus, historically, the species was distributed to various Mediterranean regions as an ornamental plant rather than a cash crop (Jashemski et al. 2002, p. 141). The Central Jordan Valley is the only place in the world where these five pioneering species of horticulture can flourish and gain economic value as cash crops.

The economic feasibility of cultivating several fruit trees simultaneously is clear. There is a parallel investment in research and development and knowledge exchange between the different species in terms of growth management techniques and propagation, making the domestication and cultivation of several species at once economically advantageous. Interestingly, all five founder fruit tree species carry the potential for vegetative propagation and, thus, are ‘preadapted’ for domestication (Fall et al. 2002). The cultivated clonal forms of each of the five founders are maintained by cutting (grape, fig and pomegranate), transplanting offshoots (dates; Figure 7) or basal knobs (olive; Zohary and Spiegel-Roy 1975). Though this study does not aim to address the question of ‘how’ fruit trees were domesticated, it is assumed that early cultivation may include the planting of seeds and stones (Fuller et al. 2023). The patterns of the change seen in stone size and shape support a protracted gradual morphological evolution for arboreal domesticates (Fuller 2018; for a different view see Zohary et al. 2012). It is possible that one of the advantages of the Central Jordan Valley compared to other regions is the relatively high ratio of seed germination success. In any event, selection processes, in terms of human action and its interaction with fruit tree genetics, are still poorly understood.

Regarding the Question of ‘when’, the date provided in this study for the crystallization of fruit-tree horticulture is 7000 cal BP (Tables 2 and S2). It renders a significant component in the establishment of the Mediterranean village economy alongside the Secondary Products Revolution (wool and dairy production and the use of animals for traction; Sherratt 1983), and not a part of state formation or urbanization (Cañellas-Boltà et al. 2018; Laneri 2018; Langgut et al. 2019), as was previously thought (Renfrew 1972). Predating urbanism by more than a millennium and a half, fruit-tree cultivation emerged primarily as a rural staple economic strategy that was only developed on a large scale during the Early Bronze Age (Fall et al. 2002; Deckers et al. 2021).

Since the crystallization of fruit tree horticulture occurred first in the Central Jordan Valley, the region enjoyed time ahead of any other region in the world with the developments that led some 1500 years later to the emergence of complex societies. It probably included the production of surplus, socio-economic changes, advances in technology, and the expansion of the exchange of knowledge, products, and raw and genetic materials. Not surprisingly, therefore, some of the earliest cities in the world, such as Bet Yerah and Jericho, were established in this core area (Figure 1).

4.2. What is the motivation behind the emergence of fruit tree horticulture?

It is still under scholarly debate what triggered the Neolithic process of the transition to agriculture and animal domestication, causing large groups of people to live close to one another and struggle with the many problems inherent in such an arrangement (Asouti and Fuller 2013 and references therein). Certainly, the decision to domesticate fruit trees ca. 7000 years ago, made the situation considerably more complex. It is likely that plantations’ longevity has had implications for future generations’ ownership of land (Abbo et al. 2015). Additionally, in contrast to annual plants, fruit trees cannot be rotated between plots, necessitating, therefore, careful premeditation and planning when land is being allocated for a fruit tree orchard (Abbo et al. 2015). Furthermore, unlike annually renewed crops, fruit tree cultivation is a long-term investment that offers relatively delayed returns since the trees require 3–8 years before bearing fruit and may not become fully productive until 10–20 years after planting (Fall et al. 2002) and it requires years of investment in research and development. All these features call for a more organized economy supported by administrations and institutions, as well as more elaborate social contracts.

Chalcolithic Tel Tsaf, located in the core area suggested in this study, is a case in point. The site provides not only one of the earliest examples of fruit tree cultivation worldwide (Langgut and Garfinkel 2022) but also an unprecedented concentration of large storage facilities (Garfinkel et al. 2009; Rosenberg et al. 2017) and evidence of administration, including the earliest stamp seal in the southern Levant (Freikman et al. 2021). The site’s findings and material culture are outstandingly rich compared to contemporary sites in the region, consisting of dense assemblages of animal bones that indicate large-scale feasts (Ben-Shlomo et al. 2010), large concentrations of ostrich eggshell beads (Garfinkel et al. 2007), a unique copper awl (Garfinkel et al. 2014), obsidian from Anatolia, Ubaid pottery from the northern Levant or Mesopotamia, Nilotic shells, and foreign beads and raw greenstone chunks (Garfinkel et al. 2007). The site clearly participated in the long-distance exchange, which must have been supported by the community’s robust economic organization, as evidenced by the extraordinary storage silos on a scale not previously unearthed in the Proto-historic Near East (Garfinkel et al. 2009; Rosenberg et al. 2017). Each building had 4–5 rounded silos, amounting to 20–30 tons storage capacity. The silos greatly exceeded the inhabitants’ needs, indicating that a complex economic system was at work, involving surplus and wealth accumulation (Garfinkel et al. 2009). The wealth of Tel Tsaf inhabitants may allow them to engage in new economic ventures like fruit tree cultivation. Fruit tree horticulture involved considerable time, labor, and gaining new knowledge without a guarantee of success, which, either way, was likely to be delayed for a long time.

Ultimately, it seems that the Tel Tsaf’s inhabitants knowingly and purposefully chose to cultivate fruit trees and were not forced into it by external pressures. For instance, climate changes characterized by fluctuations in humidity and CO₂ in the atmosphere (Richerson et al. 2001) or a decrease in seasonal contrast in the Levant (Langgut et al. 2021), were suggested as possible catalysts for the transition to agriculture during the Pleistocene/Holocene boundary. Climate reconstruction for the Levantine Mediterranean region around 7000 years ago points to favorable climate conditions (Bar-Matthews and Ayalon 2011; Keinan et al. 2019). Moreover, the proximity of the core area to the Jordan River (Figure 1) implies that water shortage was not an issue or a limiting factor either. The Central Jordan Valley is located in a transitional zone between the Mediterranean and desert environments (Figure 1a), a zone that is sufficient for sustaining rain-fed agriculture. When coupled with irrigation from the Jordan River (as in modern times), the area can be considered a productive agricultural land. The region is characterized by frost-free winters with a mean annual temperature greater than 20̊C (Srebro and Soffer 2011), permitting year-round agriculture. Assembling a package of fruit trees was, therefore, an intentional choice performed ca. 7000 years ago by the inhabitants of the Central Jordan River, who were not struggling to survive but, on the contrary, were a wealthy community where water and other climate parameters were not a limiting factor. Furthermore, this study claims that only a rich society with food surpluses and no environmental stress of any kind could allow itself to be involved in enterprises characterized by long-term investments that offer delayed returns.

Other motivations for the domestication of fruit trees may be related to the improvement of diet and food security. Fruits provided better nutrition, as the traditional cereals and legumes diet was augmented by complementary ingredients such as oil and sugar. Moreover, the fruit products are suitable for prolonged storage, guaranteeing a long-term and stable supply of these products. That includes table olives, olive oil, raisins, dried dates and figs, and date sirup (called silan/dibs). Alcohol and vinegar can be produced not only from grapes but also from pomegranates and dates. All of these long-shelf products are highly suitable for storage and long-distance trade. It can, therefore, be suggested that the domestication of fruit trees provided food security in terms of reliability, availability, stability, accessibility, predictability, and the formation of surpluses. It is reasonable to assume that these long-shelf-life products were not only suitable for long-distance trade but also for taxation. Undoubtedly, while the emergence of fruit-tree horticulture and the development of their products opened new avenues for profit accumulation, they also required new and complex economic, administrative, and political regulatory systems: New industries had to be established, central storage facilities had to be constructed, and distribution centers had to be launched. Such a complex and pioneering enterprise most probably necessitated charismatic leadership with a vision and a willingness to take risks.

Notably, Mayshar et al. (2022) made a similar suggestion regarding the Neolithic Revolution in the Near East. They argued that societies occupying areas more suitable for growing cereals are more likely to develop some sort of governance over time. The reason for this is cereals’ long shelf-life, which can make way for taxation, contrary to roots and tubers, such as potatoes, sweet potatoes and cassava, whose short shelf-life renders taxation implausible.

4.3. What is new in the core theory?

Compared to Zohary and Spiegel-Roy (1975), who argued for olive and date domestication 6000 years ago on the grounds of findings from Tuleilat Ghassul alone (Figure 1a; Zohary and Spiegel-Roy 1975; Weiss 2015), the current paper enjoyed many benefits that have accumulated over the past five decades: new discoveries and data, improvement in archaeobotanical techniques and dating methods, more regularly and systematically collection of archaeobotanical remains and advanced molecular biology and ancient D.N.A. analyses (Lancelotti and Madella 2023; Fuks et al. 2024). Thus, the ¹⁴C dates from Tel Tsaf (Tables 1 and S2) push the emergence of fruit-tree horticulture by a millennium back. The recovery of 7000-year-old convincing evidence for olive horticulture in the Central Jordan Valley indicates that the process of domestication commenced earlier. The large quantities of olive wood remains beyond the wild stock natural distribution found at Tel Tsaf probably indicates that the olive was first domesticated along the Carmel coast and the Galilee sometime during the second half of the 8^th millennium BP and that the necessary knowledge and genetic materials arrived at the Central Jordan Valley by the very beginning of the 7^th millennium BP. In turn, this adoption of olive cultivation in the Central Jordan Valley may have promoted the cultivation of the other four founder trees, which naturally occur in this region: the date palm, fig, grape, and pomegranate. Evidence for the wild presence of the latter within the core area is provided for the first time in this study (Figure 8).

Also presented in this paper is an economic logic for simultaneously cultivating several fruit trees. Undoubtedly it promotes a more efficient learning process and knowledge flow across the species, laying the foundation for the core theory. I also suggest that natural forces, such as climate deterioration, did not play any role, and the process of fruit tree domestication was most probably driven by intrinsic motivations.

The benefits of choosing the Central Jordan Valley as the core area are clear: the proximity to the Jordan River, its fertile lands, its ecological conditions (allowing for example the relatively high success of seed germination/catchment of offshoots and the high temperatures required for the ripening of dates), the early arrival of the domesticated olive and the overlap of wild distribution of the other four founders. Still, another advantage of this region emerges from a long-term historical perspective: Since proto-historical periods and till modern times, this area has been considered a more politically stable region in comparison to other nearby south Levantine regions such as the coastal plain, the Shephelah, Judea, Samaria, etc., allowing the development and establishment of new economic ventures together with their accompanied technologies.

Identifying the geographical location of the first domestication of our fruit crops is important because it allows us to gain vital information: The wild, pre-cultivated species were the most resilient and naturally adapted to constant environmental changes. When we cultivate, we trade part of that natural resilience for characteristics we want, such as higher yield. Currently, abundant fruit crops are threatened by the loss of diverse legacy cultivars, which are being replaced by a limited set of high-yielding ones (Brunson and Reich 2019; Meiri and Bar-Oz 2024). While archaeobotany traces the fruit trees’ ancestors by revealing the areas of first domestication, paleogenomics provides techniques to study the genetic diversity of landrace fruit crops. Identifying valuable genes for traits such as disease resistance, drought tolerance, and nutritional content can lead to enhanced food security and agricultural sustainability under different environmental conditions (Raggi et al. 2022; Singh and Behera 2022; Meiri and Bar-Oz 2024).

5. Summary

This study defines the Central Jordan Valley as a core area where a fruit-tree package was first assembled. ¹⁴C dates provide this irreversible event with the age of ca. 7000 cal. Years BP. The yields of the five founder fruit-tree species—olive, grape, date palm, fig, and pomegranate—provided the Central Jordan Valley’s inhabitants with better nutrition and food security. Significantly, all founder trees produce products that are highly storable and transportable and that are extremely well suited to long-distance trade, and taxation. The millennium and a half between the crystallization of fruit tree horticulture and the beginning of urbanism (ca. 7000–5500 cal. BP) enabled the formation of new industries, storage facilities, trade routes, long-distance connections, governments, and administrative systems. It is no surprise, therefore, that some of the oldest cities in the world, such as Jericho and Bet Yerah, were established in this core area.

The importance of identifying the geographical locations where our fruit crops were first domesticated by reviewing all currently available archaeobotanical information, as conducted in this study, is clear, given the immediate need to adapt horticultural practices to global climate changes and to our degraded environment. The research may promote genomic studies to identify genes for disease resistance, drought tolerance, and nutritional content, which can enhance food security and agricultural sustainability.

Acknowledgments

I would like to thank several colleagues for the exchange of thoughts and ideas: Y. Garfinkel, R. Greenberg, Y. Gadot, E. Weiss, G. Sharon, A. Langgut, A. Sasi, and E. Kremer. The study was financially supported by generous funding provided by the School of Jewish Studies and Archaeology, Tel Aviv University. A. Crivellaro, E. Asouti, M. Cavanagh, and K. Deckers are acknowledged for their assistance in verifying the wood anatomical structure of Punica granatum. Thanks are also due to N. Rosenfeld and M. Cavanagh for their help with the preparation of the figures. T. Langgut, E. Galili, S. Flit, V. Epstein, and I. Fridman are acknowledged for their assistance with photography. In addition, I would like to thank the editors and the three anonymous reviewers for their careful reading and insightful comments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All data generated or analyzed during this study are included in this article (Tables 1 and Supplementary Material 1).

Additional information

Notes on contributors

Dafna Langgut

Dafna Langgut is an Associate Professor in the Department of Archaeology and Ancient Near Eastern Cultures at Tel Aviv University (TAU), Israel. Twelve years ago, Langgut established The Laboratory of Archaeobotany and Ancient Environments at TAU. She specializes in studying past vegetation and climate based on the identification of botanical remains. She has vast paleoclimatological and paleoenvironmental experience and is involved in archaeological and environmental projects across the southern Levant, covering all times from the Lower Paleolithic to the Medieval periods. Over the years, Langgut’s environmental reconstructions focused on critical stages in human history to explore how climate affects social mechanisms and cultural processes. In this vein, for example, she evaluated possible links between environmental changes and the spread of humans out of Africa. Langgut’s research interests also include the field of botanical archaeology, which involves the collection and identification of botanical remains from archaeological contexts for the reconstruction of: ancient agriculture, plant usage (construction, fuel, diet, medicinal and cultic purposes), trade patterns etc. Under this frame, she established in Israel a new research field – the reconstruction of ancient gardens. Langgut is also the curator of pollen and archaeobotanical collections at the Steinhardt Museum of Natural History (TAU). She won the Bruno Award for outstanding and novel research in 2023.

Notes

1 Dedicated to the memory of the palynologist and anthracologist Uri Baruch, who passed away recently.