Introduction
The Tunguska event
The Tunguska event (TE; N 60°55′, E 101°57′), occurred in the early morning of June 30th 1908. A powerful explosion over the basin of the Podkamennaya Tunguska River (Central Siberia) devastated more than 2000 km2 of Siberian taiga. About eighty millions trees were knocked down by pressure pulses, a large area of trees and bushes were burnt down within the area containing collapsed trees. A number of eyewitnesses described the flight of a ‘fire ball, bright as the sun’. Seismic and pressure waves were recorded in number of observatories throughout the world. Bright nights were observed over Northern Europe and Central Asia. These different phenomena, initially considered non-correlated, were subsequently linked together as different aspects of the TE. This event was investigated by several scientific expeditions since 1927 (Kulik Reference Kulik1939). Most of the expeditions and research carried out until recent times concerned the physics and the nature of the cosmic body, whose explosion occurred at 6–8 km altitude but, despite these efforts, a detailed account of this event has not yet been reached. Tunguska-like objects may be parts of comets and thus come from distant regions of our Solar System (Bunch et al. Reference Bunch2012; Gasperini et al. Reference Gasperini, Cocchi, Stanghellini, Del Bianco, Serrazanetti and Carmisciano2012) and can be monitored only using future spaced based telescopes (Moseley et al. Reference Moseley and Mather2004; Li et al. Reference Li, Chiao, Jamieson, Faraone and Dzurak2005). There may have been similar smaller impacts that may fit more or less with the TE (Fig. 1); however, the TE is truly unique. The effects of the explosion of the Tunguska Cosmic Body (TCB) on the forest have been described by many authors, and their data gave information on the released energy and the height of the explosion. The map of the fallen trees was recently reconstructed and enlarged by Longo et al. (Reference Longo, Di Martino, Andreev, Anfinogenov, Budaeva, Kovrigin and Victor2005) using aerial surveys and data collected during investigations performed on the ground (Longo & Di Martino Reference Longo and Di Martino2002, Reference Longo and Di Martino2003). This map suggests an azimuth trajectory a = 110 ± 5°. To search for traces of the TCB, a scientific expedition was carried out by an Italian–Russian group of scientists on 1999 (the ‘Tunguska99’ expedition); see Di Martino & Longo, (Reference Di Martino, Longo, Manara and Dotto2000). The main task of the expedition was to perform geophysical investigations and to collect core samples from the bottom of the Cheko lake, a small lake ~350 m in diameter and 53 m deep, located about ~7 km NNW from the TCB explosion epicentre. This lake has a suspicious shape (it is elongated along the direction of the flight of Tunguska object), somewhat large depth for the floodplain where it is located, and it is superimposed on the meandering Kimchu River flowing from East to West (Fig. 2) at elevation 390 m just South of the 440 m hill made out of tertiary intrusion. Aerial images and digital terrain models performed during the ‘Tunguska99’ expedition show that the lake is located within an alluvial plain covered by the quaternary sedimentary deposits of the Kimchu River, that enters the lake on its SW side, and outflows ~200 m away towards North on the same side (see Fig. 2). The eastern shore of the lake is partially bounded by the SW flank of a hill (440 m of altitude, 50 m above the lake level) made of intrusive rocks (see Fig. 3), part of the diffuse Tertiary volcanic deposits. In Fig. 3 the red zigzag line indicates the extent of the damage due to the TE. No traces of the cosmic body have been found in the lake bottom sediments, but the preliminary conclusion of these investigations was that Cheko lake was formed by the impact of an asteroid fragment, which mass has been estimated in about 1.5 × 106 kg (Gasperini et al. Reference Gasperini, Alvisi, Biasini, Bonatti, Longo, Pipan, Ravaioli and Serra2007, Reference Gasperini, Bonatti and Longo2008, Reference Gasperini, Bonatti, Albertazzi, Forlani, Accorsi, Longo, Ravaioli, Alvisi, Polonia and Sacchetti2009, Reference Gasperini, Cocchi, Stanghellini, Del Bianco, Serrazanetti and Carmisciano2012). The peculiar morphology of the Cheko lake is compatible with the hypothesis that this fragment, hitting the permafrost, has triggered the formation of a ‘collapsed’ crater, subsequently filled by river water.
Fig. 1. Comparison between impact craters and nuclear explosion crater. (a) The Carancas meteorite crater (Peru). (b) The Sedan crater in the Nevada Test Site (Photo courtesy of the US National Nuclear Security administration/Nevada Site Office). (c) Scheme of an underground nuclear explosion (Courtesy of the US Department of the Navy).
Fig. 2. Satellite map of sampled trees – Tunguska area.
Fig. 3. Geological map of the area Surrounding Cheko lake.
Collapse crater
The characteristics of cosmic body impact craters are highly dependent on their final speed in the atmosphere and on the nature of the target. For a rocky target (‘hard’ impact), the maximal energy is emitted on the earth surface and is propagated at great distances causing intense devastations. The crater formed in this way is called an ‘impact’ crater and has a clear rim. For a soft target (‘soft’ impact), e.g. the Tunguska permafrost, the Carancas mud and water (Peru, see Fig. 1), the Sikhote–Alin snow and permafrost, the impacting low-speed bodies make holes in the target surface with a minimal energy emission. The impacting body losses its energy below the surface with a maximal energy emission at a great depth. The energy emitted is trapped in the target matter and cannot propagate to the neighbourhood where air-blast effects are practically absent. The material scattered around the impact point is scarce and is insufficient to form a significant rim. For a low-angle trajectory of the incoming body, some material can be deposited on the opposite side of its line of arrival. The energy emission at some depth below the surface causes the collapse of the upper layers of the target and the formation of a crater. Here, we call such a crater a ‘collapse’ crater to avoid confusion with the previously mentioned ‘impact’ craters. Recently (15 September 2007), stony meteorite created a collapse crater (Fig. 1(a)) at Carancas, in Peru. The soft impact of the meteorite on the water-saturated soil, covered by Lake Titicaca during the ice age, explains the main characteristics of the crater, 13.5 m in diameter and 5 m deep. From the crater, immediately filled by water, large chaotic chunks of soil were ejected on the crater's northern side. The impacting body was an ordinary chondrite probably of about one meter in diameter. Stony meteorites tend not to be too robust and often fragment during their atmospheric flight, giving a scatter ellipse of material on the ground. Many characteristics of a soft impact followed by the formation of a collapse crater are similar to those of an underground nuclear explosion in which the emitted energy is mainly transferred to the ground at a great depth and causes limited air-blast effects in the surroundings. As an example, we show (Fig. 1(b)) the Sedan Crater, formed when a 100 kilotonnnes explosive buried under 194 meters of desert alluvium, containing 7% of water, was fired at the Nevada Test Site on 6 July 1962, displacing 12 million tonnes of material. The crater is 98 m deep, 390 m in diameter and does not have a significant rim. Glasstone & Dolan (Reference Glasstone and Dolan1977) state that ‘if the soil is saturated and the high water table is maintained after the detonation, the crater dimensions will change with time. Slumping of the crater sides will continue until a stable condition exists for the material’. Similarly, the US Department of the Navy has estimated that dropping from the air a 1 Megaton weapon that penetrates underground in sandy soil to a depth of 15 m before exploding; the resulting crater would be about 91 m deep and nearly 427 m in diameter. The Navfac Command (1978) estimates that ‘the fraction of the energy imparted to the air in the form of blast depends primarily on the depth of burst for the given total energy yield. The greater the depth of burst, the smaller, in general, will be the portion of shock energy that escapes into the air. If the explosion is fully contained, no blast wave will form’. This is the case, schematically shown in Fig. 1(c) (Navfac Command 1978), where an explosion at a depth of ~213 m generated a collapse crater ~23 m deep, 149 m in diameter and devoid of any ring. In recent years, growing evidences have been collected in favour of the hypothesis that the formation of Cheko lake is connected to the TE (Gasperini et al. Reference Gasperini, Alvisi, Biasini, Bonatti, Longo, Pipan, Ravaioli and Serra2007, Reference Gasperini, Bonatti and Longo2008, Reference Gasperini, Bonatti, Albertazzi, Forlani, Accorsi, Longo, Ravaioli, Alvisi, Polonia and Sacchetti2009, Reference Gasperini, Cocchi, Stanghellini, Del Bianco, Serrazanetti and Carmisciano2012). Gasperini et al. proposed that the Tunguska bolide underwent fragmentation and one of its small fragments created a collapse crater now filled by Cheko lake, while the disintegration in the atmosphere of the main body was the cause of the forest devastations. In this picture, the fragment of smaller mass, piercing the 25–30 m thick permafrost layer, has triggered a massive release of H2O vapour, CH4 and CO2 that has modified the crater's dimension and geometry. Reworking and collapse of the original pre-TE riverbed may be responsible for the bean-shaped surface of the lake, while its funnel-like morphology from 5 m below the water-level, is typical for collapse craters like those in Fig. 1. Subsequently, it was suggested (Longo et al. Reference Longo, Gasperini, Bonatti, Serra and Stanghellini.2011), that the Tunguska bolide could be one of the 15–20% of Near Earth Asteroids (NEAs) that are double bodies. These NEAs, known as binary asteroids, consist of two bodies that revolve around each other at close range. The minor of these two bodies may have acted as a small fragment in the Gasperini et al. description. We conducted a dendrochronological study on trees living close to the Cheko lake and in the forest trying to establish any connection with the 1908 TCB impact.
Previous dendrochronological research
In 1991, a dendrochronological research in the Tunguska area found five spruces and one pine, that survived the TE, and made it possible to recognize a first list of 14 atomic elements as probable constituents of the TCB (Longo et al. Reference Longo, Serra, Cecchini and Galli1994, Reference Longo1996; Serra et al. Reference Serra, Cecchini, Galli and Longo1994). Subsequent work on a living spruce (Picea schrenkiana), located 7.3 km NE from the explosion epicentre, showed anomalies in tree rings at the time of the explosion (Yonenobu & Takenaka Reference Yonenobu and Takenaka1998). They identified 4 years of growth suppression after the event (1909–1912), followed by a sudden growth increase since 1913. Nesvetajlo (Reference Nesvetajlo1998) made a dendrochronological analysis on standing dead trees (he referred to them as to ‘telegraph poles’) within the area surrounding the epicentre. These were the trees that died during or shortly after the event with a dry trunk and stripped branches. No increase of C14 was detected in carbon of trees within the epicentre area (Longo et al. Reference Longo, Serra, Cecchini and Galli1994), compared to trees much further (about 700 km – Tomsk region), indicating that no extra radiation took place. The study by Vaganov et al. (Reference Vaganov, Hughes, Silkin and Nesvetailo2004) on Tunguska included 12 trees (larches, pines and spruces) living close to the epicentre area (5–7 km). Out of these samples, all survived TE, the ring growth decreased for 4–5 years after the event, and the greatest suppression was recorded in 1909. The 1908 anatomical ring structure have showed the so-called ‘light ring’, with typical lack of latewood and disrupted tracheids (strongly deformed). Deformed tracheids during the TE time seem to be related to partial or full defoliation (Vaganov & Terskov Reference Vaganov and Terskov1977; Fritts & Swetnam Reference Fritts and Swetnam1989, Savidge Reference Savidge1996; Schweingruber Reference Schweingruber1996). A similar effect was found on trees affected by the Chernobyl nuclear accident (Mousaev Reference Mousaev1996; Longo & Serra Reference Longo and Serra2006). The evidence of light rings in the Tunguska area were detected by CAT (Computerized Axial Tomography) (Longo & Serra Reference Longo and Serra2006). Here we present the results of dendrochronological analysis on 23 tree samples, belonging to the species of larch (Larix sibirica) and spruce (Picea obovata), taken at a distance between about 3.5 and 8.0 km NE from the TE epicentre (60°53′09″N, 101°53′47″E) and collected during the 1999 expedition, mostly in the area near the Cheko lake (60°57′51″N, 101°51′36″E) and close to Kimchu River. This lake has a peculiar shape and is rather deep (~53 m) to be an oxbow lake due to a cutoff of a meander of the Kimchu river. Gasperini et al. (Reference Gasperini, Alvisi, Biasini, Bonatti, Longo, Pipan, Ravaioli and Serra2007) suggested that Cheko lake was a possible impact crater produced by the fall of a fragment of the TCB. The aim of our research is to use dendrochronology of the trees living nearby Cheko lake and forest trees between the epicentre and the lake. Several of these trees survived and germinated after the TE.
Materials and methods
Investigated area
Half of the tree samples came from the area near Cheko lake (N 60°57′50′ – E 101°51′32′, altitude 354 m). This lake is located along the alluvial plain of Kimchu River, a mature meandering river flowing towards North. The lake has an elliptical shape, with the longer axis reaching 690 m directed NE–SW and the minor axis of about 350 m; the depth is about 5 m in the SW side, while it increase towards NW down to ~53 m. The placement of this lake, overlying the east side of Kimchu River meander, suggests a Holocene age. There is another, smaller lake about 2 km down the river that also has similar characteristic, having the semi-major axis more or less parallel with the trajectory of Tunguska body. Perhaps there was a smaller fragment that made that lake too (see Fig. 2). Kimchu River is a mature meandering river with several oxbow-lakes along its course. Oxbow lakes have similar depth as the river itself and once formed tend to be soon filled by sediments. Cheko lake has a depth ~53 m, ten times larger than the Kimchu local river depth, according to the field survey of 1999 (Fig. 2). Tunguska area is geologically placed into the western part of the Siberian craton made by high-grade metamorphic complexes and granites (Gladkochub et al. Reference Gladkochub, Pisarevsky, Donskaya, Natapov, Mazukabzov, Stanevich and Sklyarov2006), whose age goes back to 2.1–1.8 Ga (Archean). The oldest sediment is covered by a thick (3–12 km) sedimentary sequence with oil and gas reservoirs, trapped into shale and carbonates horizons (1 Ga – 635 Ma) overlaid by evaporatites from Cambrian (635–488 Ma) (Svensen et al. Reference Svensen, Planke, Polozov, Schmidbauer, Corfu, Podladchikov and Jamtveit2008). At the end of Permian till Triassic (252–250 Ma) the Tunguska basin was intruded by the volcanic Siberian Traps, composed by sills, dykes, followed by several millions of square kilometres thick lava flows (Kamo et al. Reference Kamo, Crowley and Bowring2006). On the way towards the surface these units made it through a thick pile of sedimentary rocks, forming sills and dykes abundant in the basin and sheets up to 350 m thick (Ulmishek Reference Ulmishek2001). Tunguska impact site is placed into the early Triassic volcanic Siberian Traps forming outcrops of sills and dykes in the hills of the TE zone. The alluvial plain with Cheko lake and River Kimchu is of Holocene age <1.7 Ka. The current geology knowledge of the area rules out the presence of carbonates and related karst topography and/or tectonic faults that could have produced the Cheko lake depression. The lake is not likely related to a recent volcanic activity because the known volcanism in this area is old (Triassic) (Kamo et al. Reference Kamo, Crowley and Bowring2006) even if it is not possible to discard the presence of a volcanic caldera or diatrema below the lake (Fig. 3). Perhaps, the Cheko lake relates to the impact of a TE fragment. Peculiar feature that support this origin is the elliptical shape, elongated NW, according to the direction of the blast wave (Gasperini et al. Reference Di Martino, Longo, Manara and Dotto2000). The results of seismic and bathymetric survey (Gasperini et al. Reference Gasperini, Cocchi, Stanghellini, Del Bianco, Serrazanetti and Carmisciano2012) have highlighted the presence of a delta-like deposit on the inflow side of Kimchu river into the Cheko lake. This suggests that Kimchu River formed a meander at the location where a fragment of the TCB, or the minor part of a binary asteroid, may have created Cheko lake.
Field samplings and dendrochronological analysis
During the expedition on summer 1999, tens of tree log section and cores were sampled. We examined 23 of them to find the effect of the TE on trees living near Cheko lake (0–100 m) and SE from it, in the direction of the TE epicentre (0–10 km), inside the forest (Fig. 2). Some of the trunk sections come from the dead trees and we checked their age using cross-dating. Four samples have shown eccentric growth due to TE blast wave that tilted young trees. Eccentricity was detectable by anomalies on ring growth and by the presence of reaction wood on the tilted side (Fig 4(a)). The evidence of deformed tracheids were recorded on several of these samples in correspondence of the 1908 ring (e.g. BO Gemelli), where it is clear an abrupt change of the tracheids alignment with the strong distortion (Fig. 4(b)).
Fig. 4. (a) Survived and tilted tree with compression wood on North side; (b) disrupted tracheids during the 1908 growing season after TE.
Tree samples were analysed using standard dendrochronological methods (Stokes & Smiley Reference Stokes and Smiley1986). The wood was sanded and measured with a sliding device (0.001 mm accuracy Velmex). Cross-correlation was made initially through ‘skeleton plots’ graphs (Fritts Reference Fritts1976), followed by tree ring widths measurement and statistical test on dating by Cofecha software (Holmes Reference Holmes1983). Each ring width curve was analysed to find sudden growth anomalies (suppression or increase) as reported in Schweingruber et al. (Reference Schweingruber, Eckstein, Serre-Bachet and Bräker1990). We converted the annual ring growth to Basal Area Increment (BAI) that is the value (mm2) of one year's basal area growth:
$${\rm BAI =} \;\pi R_t^2 \;{\rm -}\; \pi R_{t{\rm -} 1}^2, $$
where R t is the stem radius at the year t; and R t−1 is the stem radius at the year t−1. BAI is the increase in a tree's cross-sectional area from year to year. It differs from radial growth increment because it represents growth in two dimensions, providing a measure of the amount of wood added to the stem. We used it to understand health, growth – trend and competition between trees in the forest. We compared the BAI curves in the present study to those of Le Blanc et al. (Reference Le Blanc, Nicholas and Zedaker1992) (Fig. 5) to point out different tree-growth features. A comparison of our BAI curves revealed important differences between trees that survived the TE and those born afterward.
Fig. 5. Four classes of BAI: (A) young tree or not dominated anymore; (B) mature tree in closed space or with strong competition; (C) mature tree; (D) decline tree (adapted from Le Blanc, 1992).
Results and discussion
BAI analysis
The results of tree ring analysis allowed the division of the samples into two main classes of trees: those survived and those born after the TE. Inside these classes were different groups according to their BAI trend (Table 1). Trees that survived to the TE have recorded different growth trend: Group 1 includes trees not suppressed that show a short (1–2 years) or longer (4–19 years) suppression since 1908 (Fig. 6(a)). Group 2 includes trees suppressed by the competition in the forest before the TE; these after 1908 show a sudden and extended growth increase (Fig. 6(b)). Only one sample (BO 2), rather young at the TE epoch, did not show any anomalous growth effect. The class of trees born after the TE belong to groups 3 and 4. Group 3 shows a vigorous growing phase until 50–60, followed by a growth influenced by competition between trees in the forest (Fig. 7(a)). The group 4 represents those trees living close to the Cheko lake shore, and show a continuous vigorous growth without signs of competition (Fig. 7(b)).We compared all samples to BAI growth type of Le Blanc et al. (Reference Le Blanc, Nicholas and Zedaker1992) for classification purpose (Table 1). Many (36%) of survived trees, after decades of suppression growth before TE, belong to class B, with a young vigorous phase followed by competition when matured (Fig. 8(a)). Most of the trees (43%) show a similar initial trend like 7a, followed by a suppression period (B + D) (Fig. 8(b)) that started in 1950 or later and which cause is not considered in this work. Few trees (21%), which belong to the oldest group, form class C; they are the dominant ones without signs of competition (Fig. 8(c)). Considering the class of trees germinated after TE, those (50%) living inside the forest have a growth type B, with several decades of young vigorous phase followed by a mature growth, suppressed by competition (Fig. 8(d)). The trees (50%) close to Cheko lake shore or nearby have prolonged young growth phase (Type A) without competition till the days of collection (Fig. 8(e)). The growth of the trees near Cheko lake that germinated after TE supports the impact genesis of Cheko lake. We have not found any survived tree with a similar trend (Type A), germinating before 1908, affected by the TE in the area away from Cheko lake. This supports the hypothesis that the formation of the lake took place during the TE.
Fig. 6. (a) Trees that survived to TE with growth suppression after 1908 (Group 1). (b) Trees that survived and suppressed before TE with growth release after 1908 (Group 2).
Fig. 7. (a) Trees germinated after TE living close to Cheko lake with continuous young growing trend (Group 3). (b)Trees germinated after TE living inside the forest with a young growing phase followed by competition (Group 4).
Fig. 8. Type of trees according to BAI from Le Blanc, 1992: (8a) survived tree group 2 (Type B), (8b) survived tree group 2 with growth decline (Type B + D), (8c) survived and mature tree group 1 (Type C), (8d) tree born after TE living in the forest (Type B), (8e) tree born after TE living close the Cheko lake (Type A).
Table 1. Sample details: Names, GPS Locations, Dating range, Group, Eccentricity, Release after 1908, Suppression after 1908, BAI type, and Notes
a Dr Kletetschka's sample.
b Eccentricity analysed only in cross-section.
Grey cell = trees born after TE.
LASI = Larix sibirica ledeb.
PCOB = Picea Obovata Ledeb.
While looking in detail at the BAI of trees that survived to 1908 TE in vicinity of the shore of Cheko lake (Fig. 9), we found trees with a suppression growth, that aged 16 and 60 years, before the disturbance of the event took place (see samples BO24, BO42 in Fig. 10(a)). The temporal length of the growth suppression reflects their age at the moment of the event. Similar BAI trend of suppressed trees, lasting tens of years before the event, was found on trees living further from the lake, inside the forest (e.g. T 11 and BO Vulfing, see Fig. 10(b)).The similarities in the growth suppression before the event probably indicate a dense forest spread throughout the region, including present extent of Cheko lake.
Fig. 9. Detailed map of sampled trees nearby Cheko lake.
Fig. 10. BAI graphics of trees living near Cheko lake.
Assuming Cheko lake already existed before 1908, we would not expect suppressed growth trend before TE, while a typical linear and/or exponential positive growth trend, as for the trees germinated after TE, have no competitors ‘group 3’ living along the shore lake (Fig. 7(b)). The lack of competitors induced vigorous growth pattern in this tree group that germinated after 1908 and growing along the shoreline (see BO7 and BO37 in Fig.10(c)). Similar trend was found just after TE in tree rings of survived trees in proximity of Cheko lake (BO 42) and several years later (BO24, see also Fig. 10(a)).
So the BAI growth trend supports the hypothesis of a neoformation of Cheko lake due to TE in 1908 during which the course of Kimku River would have changed its position in respect to the new formed basin. Further dating and BAI analysis of trees living along the shoreline lake and its surroundings will provide more details of the events that took place during this event (not part of this paper).
Trees’ tilting
In addition to BAI results, we detected the effect of TE from the tree trunk morphology on the 16 cross-section samples. Many of the survived trees did not show eccentric growth, except four of them near the Cheko lake, that were tilted towards N (or NE, NW) consistently with the blast wave direction coming from E 110○ (Fig. 4). This suggests that these four trees were tilted after the formation of collapsed crater in the area of today's Cheko lake. Some trees could be sheltered by local topography from TE explosion blast wave, because these did not present ring grow anomaly even if very close to the epicentre. The direction of the tilted trees agrees with the updated map of fallen trees (Longo et al. Reference Longo, Di Martino, Andreev, Anfinogenov, Budaeva, Kovrigin and Victor2005).
Discussion
We investigated the effects of TE on trees living within the impact zone, an area of about 2000 km2. We found TE effects on trees both in morphology (trees’ tilting) and in wood anomalies as sudden growth increase and/or decrease. The use of BAI, rather than ring width, allows a more comprehensive view (Nesvetajlo Reference Nesvetajlo1998; Yonenobu & Takenaka Reference Yonenobu and Takenaka1998; Vaganov et al. Reference Vaganov, Hughes, Silkin and Nesvetailo2004; Fantucci et al. Reference Fantucci, Di Martino and Serra2012) of the effect on tree growth and the variation of competition before and after the TE. Most of the previous work (Nesvetajlo Reference Nesvetajlo1998; Yonenobu & Takenaka Reference Yonenobu and Takenaka1998; Vaganov et al. Reference Vaganov, Hughes, Silkin and Nesvetailo2004) detected growth stress from observing tree ring decrease lasting few years, sometimes followed by a sudden growth increase. Instead, our work suggests an existence of both suppression and increase of the growth after the TE. For the suppression case we detected a growth reduction, starting in 1908 or 1909, and lasting from a mean of 3 years up to 14 years, followed by a growth increase lasting 4–14 years (Fig. 6(a)). Trees that survived the TE have shown an accelerated growth. These trees had suppressed growth in the forest before the TE and after that, with the death of their neighbours, suddenly appeared in better living conditions without competition for the light and nutrients (Fig. 8(a) and (b)). Nesvetajlo (Reference Nesvetajlo1998) observed similar growth trend close to the epicentral area. In addition to the survived trees, we analysed trees born after the TE, living inside the forest and near the Cheko lake shore. The oldest ones (33%) were born in 1923 (15 years after the TE). BAI analysis of trees, born after the TE, living near Cheko lake shore and within the forest has shown two different growth trends. Trees near the lake shore have continuous young vigorous growth (Type A – Le Blanc) until 1999 (the day of collection) (Fig. 8(e)). Trees living inside the forest, instead, have a trend with an initial vigorous growth followed, after 10–30 years, by a more suppressed trend, indicating a re-established competition (Type B – Le Blanc) (Fig. 8(d)). The difference in the growth rate between the trees born after the TE inside the forest and near the Cheko lake shore point out that they were living in a different environment. Near the lake trees do not need to compete one with each other, while inside the forest they have to. The dendrochronological analysis results suggests a recent age of the Cheko lake, and supports the hypothesis that the lake may have been formed by a TBC fragment or a smaller component of a binary asteroid impacting on permafrost and the soft alluvial deposit of the Kimchu river plain.
The last but not least point is that the TE took place at an altitude between 5 and 10 km (Foot & Yoon Reference Foot and Yoon2002). This is consistent with the distance of Cheko lake from the epicentre of the explosion as well as a similar recent event in Chelyabinsk, where the largest part of cosmic body flew from an altitude exceeding 20 km comparable horizontal distance prior to an impact into the Chebarkul lake (Borovicka et al. Reference Borovicka, Spurny, Brown, Wiegert, Kalenda, Clark and Shrbeny2013).
Conclusion
In this study, the growth comparison between trees that survived to the TE and those born after, analysed by BAI, led to a new direction of research: the state of competition between the trees. Many of the survived trees had a growth suppressed before the TE by the presence of the competing trees, and a sudden and strong increase for 10–30 years, before a new competition arises. Six trees have shown growth suppression, generally for a few years (1–5). A comparison between trees born after the TE close to the Cheko lake and those inside the forest have shown an absence of competition between trees living close the lake shore, while the others, after 10–30 years, suffered for the competition of the neighbours. The oldest tree born after the TE goes back to 1923, 15 years after the event, the time lag of resilience displayed by this research. Trees that recorded the presence of the lake before 1908 were not found; all analysed trees near the lake shore show a typical young growing phase without competition, different from those inside the forest that after a young phase show the sign of competition on the BAI curve. In summary, we do not claim that we proved that the Cheko lake is of impact origin. However, our analysis of trees around the lake and near the impact epicentre show that the tree record from this area is consistent with the impact hypothesis where the soft impact of a TCB fragment or of a component of a binary asteroid could have originated the Cheko lake.
Acknowledgements
We thank the Fondazione Cassa di Risparmio di Torino for financial support. The work was supported by MEYS (LK21303, MSM0021620855 and research plan RVO:67985831 for GK,). We thank Giuseppe Longo for his precious suggestions for the article, Kristyna Cizkova for help with digitizing the geology map of the area. Following people helped with field work and sample preparations: Karel Kletetschka, Andrea Kletetschkova, Vilem Mikula and Ladislav Nabelek.
