1 Introduction

1.1 The neolithic revolution fire problem

The Levant is one of the main centers of domestication of wild plants for human use during the early Holocene Neolithic revolution, evidencing the first transition to agriculture (Wood 2011). Many causes for this revolution have been proposed (e.g. Barker 2006; Bar-Yosef and Belfer-Cohen 1992; Cappers et al. 2002; Colledge et al. 2004), including a short period of rapid climate instability (Borrell et al. 2015). While the emergence of agriculture was associated with environmental change, it also enhanced the capability of newly established human settlements to modify the environment with various techniques, including fire (Blockley and Pinhasi 2011).

The systematic analysis of patterns of wildfire occurrences – fire regimes – shows that the Levant, with its biomass and hot, dry climate, has always been a hot spot of fires (Conedera et al. 2009; von der Kammer and Ignatieff 2024). Natural fire regimes have changed markedly during the Neolithic period (Turner et al. 2010), and charcoal records seem to have peaked in a global scale (Power et al. 2008). Fires during the early Holocene were associated with various natural elements, such as climate forcing, weather patterns, availability of combustible material and ignition sources (Vannière et al. 2008, 2011). However, fires also represent a major and ancient method of anthropogenic environmental control (Keeley et al. 2011). Therefore, fire regimes in the Levant could both influence and be influenced by the Neolithic agricultural revolution.

Hominins of the southern Levant possessed fire as a maintainable technology at least since the early middle Paleolithic (~ 300 ka, Karkanas et al. 2007), and possibly much earlier (Alperson-Afil 2008). An early study by Lewis (1972) used ethnographic comparisons to suggest that burning served as a landscape management technique toward plant domestication in the Levant, and carefully controlled fire usage enhanced the cultivation of herbaceous plants. More broadly, it is known that wildfires help to maintain open grasslands by promoting their growth while hindering tree encroachment. Wild ancestors of domesticated plants could thrive following low-intensity fires, which are common in Mediterranean ecosystems during the dry season (Zohary et al. 2012). Naveh and Carmel (2003) suggested that during the Neolithic revolution, what was earlier a mutually beneficial relationship between humans and nature, shifted to human control, leading to practices such as controlled burning. During the pre-pottery (or early) Neolithic period (Between 8500 − 7600 BCE), people began changing the land from areas covered in year-round green plants to fields of seasonal winter grasses by repeatedly burning the land, which encouraged new grass growth after each fire (Matthews 2016; Miebach et al. 2022). However, fires can also harm large-seeded annual grasses, as their fire-tolerance depends on the species (Blumler 1991). In spite of the importance of fires in the Neolithic revolution, no study has clarified the relative importance of natural vs. anthropogenic fires on a regional scale.

Here an early Holocene charcoal record is compared with other records of the same period, in order to gain further insight into the cause of the early Holocene fire regime and associated loss of soil and vegetation. The possible impact of these changes on early agriculture is discussed.

1.2 Environmental setting

The studied records and sites (Fig. 1) represent various environments in the southern Levant, from hilly to flat base-level settings and from Mediterranean to arid climates. Most of them act as various types of traps, whose deposits provide valuable records of early Holocene environmental change.

Fig. 1
figure 1

Location map. 1 – Hula basin; 2 – Sha’ar Hagolan; 3 – Munhata; 4 – Mekhora; 5 – Wadi el Ahmar; 6 – Gilgal and Netiv Hagdud; 7 – Jericho; 8 – Ramat Rahel; 9 – Har Nof (Jerusalem) Cave; 10 – Motza; 11 – Soreq Cave

Full size image

Fire events cause sudden increases in charcoal fluxes that move from the fire site through air and water transport systems, eventually entering aquatic deposits. Creating an accurate fire history record depends on identifying these charcoal records within lake sediments. The analysis of micro-charcoal and other combustion products deposited in sediment traps has become a well-established method for reconstructing local, regional, and even global histories of biomass burning (Leys et al. 2017; Marlon 2020; Patterson et al. 1987).

Bodies of water and marshlands have existed within the pull-apart Hula basin of the Jordan Valley, dammed by its southern fault and basalt flows (Heimann et al. 2009). Before the shallow Hula lake (65 m asl, Fig. 1) was artificially drained in the late 1950s, it covered few square kilometers, with an equivalent area of wetland further to the north. The Hula Valley borders to the west by the Naftali limestone and dolomite mountains of Upper Galilee (up to 1200 m asl) and to the east by the volcanic Golan Heights (up to 1300 m asl). The climate of the region is characterized by cool, rainy winters and hot, dry summers. The mean annual temperature in the valley is 21 °C, reaching a maximum of 40 °C in the summer, while in the adjoining mountains, the mean annual temperature drops to 16 °C. Annual rainfall in the valley is 500 and 700 mm yr− 1 in the south and north respectively, increasing to ~ 1000 mm yr− 1 in neighboring mountains. During the Quaternary the Hula basin has acted as a trap for pollen and micro-charcoal (Weinstein-Evron 1983).

Between the Sea of Galilee and the Dead Sea, the lower Jordan Valley was relatively densely occupied during the Neolithic period (Fig. 1). The lower Jordan Valley is climatically diverse, stretching from the Mediterranean climate of Sea of Galilee to the hyper-arid climate of the Dead Sea (Frumkin and Shtober-Zisu 2024). The valley bottom is covered mostly by the sediments of Lake Lisan, the glacial period predecessor of the Dead Sea, and overlying sediments. Following the lake-level fall, the Jordan River flowed from the Sea of Galilee to the Dead Sea along the previous lake-level bed, forming an elongated depression with the Jordan floodplain at the bottom. The Jordan River is meandering along this sub-valley. Since the anthropogenic extraction of upstream waters during the 20th century, the river stopped inundating the floodplain, and its route has stabilized. Springs and flowing water are common along some parts of the lower Jordan Valley. Some of these were associated with Neolithic settlements and were affected by the major change in hydrogeologic configuration since the last glacial period (Levy et al. 2020).

The backbone hills, 40–60 km from the east Mediterranean coastline, comprise a thick sequence of carbonate rocks of late Cretaceous to Eocene age. The present water-divide region, lying between the Mediterranean and Jordan Valley base levels, has been exposed to karstification during the late Cenozoic (Frumkin 1993). Rising today up to ~ 1000 m asl, and experiencing 500–600 mm yr− 1 precipitation, the region has speleothem-decorated caves, combined with other karst features (Frumkin 2024). Rising from the west, the rain clouds coming from the eastern Mediterranean are adiabatically cooled by the hills, producing orographic precipitation. On the eastern, rain shadow side of the ridge, the air undergoes compressional warming and drying, producing a local desert (Vaks et al. 2003). The temperatures on the eastern flanks are warmer by several degrees than those on the western side.

The speleothems discussed here are from Soreq and Har Nof (Jerusalem) caves (at 400 and 730 m asl, respectively), on the western side of the backbone hills, experiencing ~ 500–600 mm yr− 1 precipitation (Bar-Matthews et al. 2017; Frumkin et al. 2022). The Har Nof (Jerusalem) cave demonstrated the best isotopic equilibrium in the region, probably due to its small size and high humidity (Fig. 2) (Frumkin et al. 1994). The speleothem chronology is based on high-accuracy U-Th dates. Two speleothem proxies are used - Strontium and Carbon isotopes, assumed to reflect the vegetation and soil cover above the caves, respectively.

Fig. 2
figure 2

Har Nof cave in Jerusalem (site 9 in Fig. 1). a The cave entrance in the road-cut which exposed it. Note the shallow soil on the surface; b The eastern wall of the cave, decorated by speleothems, as seen by internal photogrammetry draped on Lidar measurements (surveyed by A. Frumkin on 2024); profile (c) and plan (d) of the cave, respectively (surveyed by Boaz Langford on 2024). The location of the studied speleothem is indicated in red

Full size image

2 The studied records

2.1 Charcoal record

A Hula lake core was used by Turner et al. (2010) to investigate the early Holocene fire history (Fig. 3a). The core is 16.25 m long, consisting of pale grey-buff calcareous clay marls, and dated by 14C. The micro-charcoal particles in the core, combined with stable isotopes and pollen data, demonstrate strong correlations between biomass burning patterns, climate, and vegetation throughout the Holocene. Climate changes appear to have primarily driven regional-scale fire histories during this period. Increased moisture and higher temperatures enhanced vegetation community productivity, resulting in greater fuel availability. In particular, a major charcoal peak started ~ 10 ka, when charcoal particles increased to ~ 300% of the normal Holocene values (Fig. 3a). Its final stage is not clear due to a hiatus (Turner et al. 2010). This unusual peak deserves further discussion concerning its cause and associated environmental impacts. The charcoal record is compared with speleothem isotopic records from Israel’s backbone hills.

Fig. 3
figure 3

Studied records of terminal Pleistocene-Holocene environment in the southern Levant. a Micro-charcoal record of the Hula basin, modified after Turner et al. (2010). b δ13C vPDB of speleothems from Soreq Cave, modified after Bar-Matthews et al. (2017). c Lake level at the Dead Sea basin, compiled from Bookman (Ken-Tor) et al., 2004; Frumkin 1997; Lisker et al. 2010; Goldsmith et al. 2023; Torfstein 2019; Migowski et al. 2006. d87Sr/86Sr ratios of speleothems in backbone hills around Jerusalem: Soreq Cave (red line and grey dots, showing the distribution), modified after Ayalon et al. (2013); Har Nof (Jerusalem) Cave (red dots and blue line), modified after Frumkin et al. 2022. e Histogram showing the distribution of OSL age measurements of old brown soils underlying agricultural terraces at Ramat Rahel, Jerusalem (modified after Davidovich et al. 2012). Green rectangle shows the period of large Neolithic farming villages over redeposited soil accumulations in valleys (Kenyon 1957; Noy et al. 1980; Bar-Yosef et al. 1991; Twiss 2007; Garfinkel et al. 2009; Khalaily et al. 2020; Delage 2007). f 33°N and 65°N summer insolation curves

Full size image

2.2 Carbon isotopes

Carbon isotopes in speleothems reflect mainly the overlying vegetation cover, including the amount of vegetation, and C3 vs. C4 plants, soil biogenic activity (e.g. McDermott 2004), as well as smaller environmental effects such as water stress, high temperature and CO2 level (e.g. Cerling and Quade 1993; Frumkin 2009; Dreybrodt and Scholz 2011; Breecker 2017). C3 plants would contribute carbon with very negative δ13C values (~ -27‰ vPDB), C4 vegetation would provide carbon with less negative δ13C values (~ -12‰), while both are buffered by higher bedrock δ13C values (around 0‰). High δ13C values may indicate bare host-rock with sparse vegetation cover.

A well-dated speleothem record from Soreq Cave (Fig. 3b) demonstrates an early Holocene sharp peak of positive δ13C, rising from 8.5 to 7.8 ka to -4‰ above the background Holocene values of ~ -13 ˗˗ -10‰ (Bar-Matthews et al. 2017). The steep topographic slopes above the cave indicate that a loss of vegetation would possibly increase soil degradation due to enhanced erosion. This scenario is further tested with Sr isotopes of speleothems (below).

2.3 Dead Sea level

As a terminal lake in an arid region, the Dead Sea serves as an excellent gauge for the hydrologic budget of its catchment, integrating both Mediterranean climate systems of the southern Levant, such as Cyprus lows, as well as southern rainfall signals (Frumkin 1997). The Dead Sea evidence indicates a dry interval ~ 8.5–7.8 ka (Goldsmith et al. 2023). Migowski et al. (2006) observed a depositional hiatus in three cores drilled along the Dead Sea shore and summarized: “The evidence from the three cores combined suggests a rapid drop of lake level ~ 8.1 cal kyr BP from above 412 to below 430 m bmsl, and rising of the lake above 430 m bmsl ca. 300 year later”. This was the most severe natural drop of Dead Sea level during the Holocene (Fig. 3c). The severe lake level drop is noted by a hiatus both in the Dead Sea and the Hula Lake (Fig. 3a). In Mount Sedom, the severe drop of Dead Sea level exposed the rocksalt of the diapir for the first time, promoting the formation of the first salt cave (Frumkin et al. 2021).

2.4 Strontium isotopes

The calcite comprising speleothems in Israel shows 87Sr/86Sr ratios typically around 0.7076–0.7087 (Frumkin and Stein 2004), with contributions of Sr from the carbonate host-rock of the caves (87Sr/86Sr ratios of ~ 0.7074-6), and from the Terra Rossa soil above the cave (87Sr/86Sr ratios of ~ 0.7087 and higher). Speleothems with relatively high 87Sr/86Sr ratios (~ 0.7085-7) indicate derivation of the Sr mainly from the Terra Rossa soil, which is enriched in dust and clay minerals. Other factors and sources also contribute to the Sr isotopic values, but to a lesser extent. When no surface soil is available and host-rock dissolution predominates, 87Sr/86Sr can be expected to approach the ~ 0.7076 value of Mesozoic carbonate rocks (Bar-Matthews et al. 2017; Frumkin et al. 2022 and references therein).

The Sr isotopes records from Soreq and Har Nof caves (Fig. 3d) demonstrate high late-Pleistocene values (~ 0.7084), followed by sharp decrease to low values (~ 0.7078–0.7079) during the early Holocene, indicating loss of soil cover. The peak is less sharp than the charcoal and carbon peaks, yet it appears across all the Mediterranean zone of the southern Levant, from Mt. Hermon in the north to the Judean backbone hills in the south (Ayalon et al. 2013; Frumkin et al. 2022). The scatter of points of the low Sr isotopes values (Fig. 3d) may indicate a gradual loss and recovery of soils, evident today above the caves (Fig. 2a). This is further investigated below by field observations.

2.5 Eroded soils and human settlements

Today, the backbone hillslopes of Israel commonly lack thick soils (Fig. 2a) (Frumkin 2024). The Holocene history of the soils on slopes is indicated by OSL dating of soils at Ramat Rahel (site 8 in Fig. 1). (Davidovich et al. 2012). The early Holocene has the smallest number of ages until modern times, indicating soil degradation, in agreement with the Sr evidence (Fig. 3d-e).

The eroded soils were usually redeposited in depressions. These include local soil pockets between bedrock outcrops of karstic karren (Fig. 4), thick accumulation of eroded soils in sub-.

Fig. 4
figure 4

Local terra rossa soil pockets in karstic karren terrain on the backbone hills near Jerusalem

Full size image

horizontal local valleys (Fig. 5a), and within regional base level valleys, such as the lower Jordan Valley (Fig. 5b). Large aggrading fans were accumulated in the valleys, with typical brown terra rossa-derived clays. These overly the white lacustrine Lisan sediments (Fig. 5b), whose top layers are dated to ~ 13 ka, associated with the final retreat of the lake (Müller et al. 2022).

Fig. 5
figure 5

Reworked terra rossa soils in the vicinity of the southern Jordan Valley. a Oblique Google Earth view from Mekhora (site 4 in Fig. 1), looking to south-west, showing brown soils redeposited in small local valleys close to the eroded slopes, and location of similar deposits at the Jordan Valley; b Close-up of redeposited soils on top of the ~ 13 ka uppermost Lisan sediments (Wadi el Ahmar, site 5 of Fig. 1). The nearby Neolithic sites of Gilgal (Noy et al. 1980) and Netiv Hagdud (Bar-Yosef et al. 1991) were constructed on these fertile redeposited brown soils

Full size image

Within the southern Jordan valley, on top of the fan sediments, some of the largest Neolithic settlements were uncovered (Twiss 2007) (Fig. 3e), including Jericho (site 7 in Fig. 1) (Kenyon 1957), Gilgal (Noy et al. 1980), Netiv Hagdud (Bar-Yosef et al. 1991) (6 in Fig. 1). At the central Jordan valley they extended into the late (Pottery) Neolithic e.g. at Sha’ar Hagolan (Garfinkel et al. 2006, 2009)(Fig. 6), Munhata (Perrot 1966), as well as smaller, or less-excavated sites (Garfinkel 1994; Garfinkel and Dag 2006; Garfinkel et al. 2020). On the other hand, within the backbone hills, where the soils were mostly eroded from the slopes, large Neolithic settlements are mainly concentrated on local thick sediment accumulations in valleys, such as Motza (site 10 in Figs. 1 and 7) (Khalaily et al. 2020). Thus the general settlement pattern indicates diminished population on rocky hillslopes, in favor of soil accumulations in valleys, where agriculture could be easily practiced.

Fig. 6
figure 6

The stratigraphy and late Neolithic activity at Sha’ar Hagolan, central Jordan Valley. a the stratigraphic section, including laminated Lisan sediments (bottom) overlain by fluvial deposits; b Neolithic well at Sha’ar Hagolan late Neolithic site during the excavation (site 2 in Fig. 1) (Garfinkel et al. 2006), radiometrically dated to ~ 8.4–8.2 ka. The well was dug into the Lisan lacustrine sediments (finely laminated) through the overlying reworked soils and gravel (courtesy Y. Garfinkel)

Full size image
Fig. 7
figure 7

The Neolithic site of Motza (site 10 in Fig. 1), showing structures built on thick redeposited soils. The high concentration of clay precludes infiltration of rain water. Har Nof Cave (Fig. 2) is located on the eroded slope of the hill in the background (site 9 in Fig. 1)

Full size image

3 Discussion

The compiled evidence (Fig. 3) indicates a major environmental collapse in the southern Levant during the early Holocene, culminating ~ 8.2 ka. This event included loss of vegetation cover, stripping of soils at hillslopes and associated deposition in valleys, and high concentration of charcoal particles, suggesting a pulse of extreme fires. Considering the dating uncertainty of the various records, it is not easy to distinguish between causes and effects. However, the loss of vegetation seems to be coeval or closely follow the fire pulse, apparently demonstrating cause and effect, respectively. A tipping point threshold was apparently crossed by the intensive fire regime, causing dramatic stripping of vegetation during the late Neolithic period (Pottery Neolithic).

The Sr isotopes record indicates, however, that the loss of soil cover and deposition in valleys were a more prolonged process, occurring since the desiccation of Lake Lisan (during the Natufian Period) and into the end of the Neolithic period. The stripping of vegetation by intense fires could have accelerated the erosion process, thus soil erosion peak was coeval with the fires and vegetation loss (Fig. 3a-d). The major drop of Dead Sea level indicates an extremely dry episode which may have promoted the fires and loss of vegetation.

The ignition of the extreme fires cannot be volcanic, as the last volcanism near the Hula basin, at the northern Golan, occurred ~ 100 ka (Weinstein et al. 2013; Shtober-Zisu et al. 2018; Frumkin and Naor 2019), and no later volcanism is known in other southern Levant areas.

An anthropogenic fire ignition is possible, but it seems improbable that humans, concentrated locally, ignited large fires for a fraction of the entire Neolithic period, causing major environmental collapse, including the loss of most vegetation and hillslope soils.

The low lake level suggests a major dry climatic event. Lightning is a major cause of natural ignition of wildfires in the Mediterranean basin (Pérez-Invernón et al. 2021), so increased dry thunderstorm intensity, in addition to dry vegetation, is the most probable scenario (e.g. Beffa et al. 2016; Burjachs and Expósito 2015). Dry thunderstorms are known to cause fires especially in dry regions (e.g. Huang Yan et al. 2015), particularly during the end of the dry season under strong seasonal climate (e.g. Meroney 2007).

Intrusion of thunderstorms from the south may indeed have occurred at the climax of the Holocene Humid Period further to the south (HHP, e.g. Dinies et al. 2015; Neugebauer et al. 2022), associated with orbital forcing: Solar radiation peaked during the HHP, causing the northward migration of low-latitude storms, with significant environmental and cultural changes in Africa and Arabia (Fig. 3f). The resulting enhanced Nile flow during the HHP is associated with deposition of sapropel S1 in the East Mediterranean (Grimm et al. 2015). The Sahara Desert belt became habitable for savanna vegetation, fauna and humans, before drying out again in the mid-Holocene (e.g. Butzer et al. 1972; and see also Enzel et al. 2015). The magnitude and geographic extent of northward penetration of this climatic event towards the southern Levant remains debated (Drake et al. 2022; Fleitmann et al. 2022; Lüning and Vahrenholt 2019; Quade et al. 2018; Torfstein 2024). Within the southern Levant, only one U-Th date was reported from the Negev tufa deposits (Waldmann et al. 2010), and no speleothems from this period were found in the Negev Desert (Vaks et al. 2006, 2010). This indicates that the early Holocene wetness intrusion into the southern Levant was weak or non-existent.

The history of arid zone lakes or wetlands in Northern Arabia, south of the Levant, may serve as proxy for HHP penetration to the nearby desert, where Mediterranean rainfall systems do not penetrate. The sparse lake records in Northern Arabia indicates that HHP penetration lasted only for few hundred years between ~ 8 and 8.6 ka (Dinies et al. 2015; Shanahan et al. 2015; Neugebauer et al. 2022). This included a two century-long peak humidity ~ 8.2 ka. While the humidity peak did not reach the southern Levant (above), the HHP was marginal while the dry Saharan belt moved northward due to high radiation. Thus the similar chronology of the fires (Fig. 3a), vegetation loss (Fig. 3b), and low Dead Sea levels (Fig. 3c) raises the possibility that dry thunderstorms did reach the southern Levant ~ 8.8 – 7.9 ka, causing the peak in fire intensity and loss of vegetation, while the water budget of the Dead Sea was negative. The brief duration of this peak is consistent with the lack of the δ13C peak in the Jerusalem record (Frumkin et al. 2000), apparently due to a short-term hiatus.

On the other hand, the HHP marginal penetration from the south was coeval with a centennial-scale cold and dry event around 8.2 ka in adjacent northern regions and across the northern hemisphere in general (e.g. Parker and Harrison 2022; and reference therein). During this event a major increase in fires was recorded in the west Mediterranean too (Davis and Stevenson 2007). Modeling simulations suggested a latitudinal gradient in the magnitude of the 8.2 ka event, with more pronounced vegetation responses to this event in northern Europe and weaker responses in North Africa (Li et al. 2019). The records from the southern Levant discussed here demonstrate however a severe to catastrophic response at the east Mediterranean region, probably due to the marginal intrusion of HHP-related dry thunderstorms.

Further insight may be gained from a previous similar event, yet more extreme, which occurred in the southern Levant during MIS 5e, according to various proxies, including the ones used here (Frumkin et al. 2022). Extremely high concentration of charcoal particles in the Dead Sea core during MIS 5e reflects frequent and intensive fires. This high charcoal concentration contrasted with background values before and after MIS 5e, which were extremely low, amounting to few percent of the MIS 5e values (Chen and Litt 2018).

While the normal δ13C values reflect the C3 vegetation common in the Har Nof cave region today, the extreme MIS 5e pulse of δ13C, amounting to -4‰, indicates a removal of C3 vegetation above the cave (Frumkin et al. 2000). Such high values cannot be attributed to C4 vegetation only, as such vegetation would result in lower δ13C values. In addition, a change of vegetation from C3 to C4 and back is expected to occur as a gradual process rather than as a sharp, almost instantaneous peak. The low 87Sr/86Sr ratios in the Har Nof and Soreq speleothems suggest severe soil and floral removal from the overlying hillslope (Bar-Matthews et al. 2017; Frumkin and Stein 2004). The most probable explanation of the extreme δ13C and 87Sr/86Sr fluctuations is strong fire regime which consumed the vegetation on the hill slopes, causing severe erosion and soil removal during MIS 5e (Frumkin et al. 2022).

An additional feature of the MIS 5e event is a major spike of fungal spores in the Dead Sea core, coeval with the charcoal spike (Chen and Litt 2018), indicating an explosive spread of these stress-tolerant and opportunistic organisms (Mudie et al. 2011). Fungi are known to be among the first colonizers after forest destruction by severe fires (Claridge et al. 2009). Fungal spores were not counted yet in the Hula Holocene core discussed here.

The general fire regime of MIS 5e is comparable with that of the early Holocene event, albeit the Holocene event had a lower magnitude. This is in agreement with additional observations of ~ 100 kyr cyclicity of peak fire events, associated with warm periods controlled by orbital forcing (Kappenberg et al. 2019).

The Neolithic revolution could be a natural consequence of human evolution and symbolic development (e.g. LePoire 2023; Shavit and Sharon 2023; Simmons 2011; Svizzero 2017; Watkins 2018; Weisdorf 2005). Yet, the coeval development of the Neolithic agricultural revolution and the concentration of large farming communities over aggraded reworked soils suggest that the natural processes discussed above had an important effect on the human populations. Natural habitats on hilly slopes were eliminated by fires and soil removal, forcing populations to change their subsistence strategy by concentrating in valleys and developing their own agricultural economy, which was the only way to support large populations. While the process of soil loss occurred across the entire Neolithic period, it culminated with the intensive fire regime ~ 8.2 ka.

4 Conclusions

The available evidence associated with the early Holocene charcoal peak suggests a natural cause for fires, namely the northward movement of climatic belts to the southern Levant. This probably included a high intensity of dry thunderstorms with lightning that ignited the vegetation at the end of the dry season, coeval with the northern hemisphere 8.2 event. The intense fires caused reduction of biomass and plant stature, including elimination of woody plants. While plants were eliminated on slopes. Savannah-like grasses could penetrate reworked soil-accumulation sites in the valleys, leading to 13C enrichment in associated deposits (e.g., Keeley and Rundel 2003; Karp et al. 2018). The intense fires increased the stoniness, surface runoff and erosion by reducing vegetative cover (Shakesby 2011). Thus the soil hydrologic properties changed, providing a readily erodible layer of sediment and charcoal (e.g. Shtober-Zisu and Wittenberg 2021). The intensified heat, dryness, and fire activity probably eliminated the C3 vegetation and prevented the establishment of woody plants. Dry grasses could readily be ignited by lightening inflicted by dry thunderstorms. This probably caused major soil removal from most hillslopes, causing lower Sr isotopic values in speleothems. The suggested scenario of dry thunderstorms and associated fires reflects atmospheric instability related to the high solar radiation and HHP marginal penetration (Zhisheng et al. 2015). The soil removal and its redeposition in valleys explains the concentration of large Neolithic sedentary communities over such depositional basins which were most convenient for developing early farming practices. Human behavior and cognitive abilities must have adapted to the extreme environmental change. This cognitive adaptation might have influenced unprecedented behaviors, such as initial penetration into deep complex caves in the backbone hills of the southern Levant (Ullman et al. 2024). Geomorphic stability could be re-established only after the end of the extreme fire regime event (e.g. Shakesby and Doerr 2006), providing for human habitation also of hilly areas, where soils gradually regenerated.