Скачать бесплатные шаблоны Joomla

M. Gurskaya. Preliminary tree-ring dating of historical wood from Ust-Voykar settlement (15th-20th Centuries), Northwestern Siberia // Tree-rings in archaeology, climatology and ecology. Vol. 4. Proceeding of Dendrosymposium, April 21-23th 2005, Freibourg, Switzerland. 2006. Vol.54. 2006. P. 236-243.


Calendar dates for archaeological monuments and environmental reconstruction in Northern Siberia are an important application of dendrochronology to the history of Siberia. Historical monuments from the northern taiga of Western Siberia represent a unique source of knowledge about people's past life in the harsh environment of the Russian North. Shiyatov and Hantemirov (2000) studied historical settlements, Nadym and Yarte IV in Western Siberia; they found that a great number of trees and other woody plants were used in their construction. Timber from these settlements was used to define building-period dates and the tree species used in construction were identified. This work revealed that long tree-ring chronologies could be developed from tree-ring series constructed from historical timbers from northern settlements. Tree-ring chronologies provide a proxy record of past climatic conditions (Kolchin 1963, Bailie 1982). However, there a limited number of long tree-ring chronologies from the northern part of Western Siberia exist. We believe that historical timbers, recovered from high latitude settlements especially those near the Russian Polar circle, are a great source of tree rings for developing long tree-ring chronologies. The subject of this paper, the Voykar settlement from North-Western Siberia, supplies one such chronology.

The Ust-Voykar settlement

The Voykar settlement, situated on the west bend of the Gornaya Ob River (65°41'N, 64°41 'E), is one of the largest and best-known settlements of the XVI - XIX centuries. There are various myths and legends attached to the historical events that happened in this settlement. It was a trade centre for the native "Ostyak" and "Samoed" tribes from the vast territory of Siberia and a centre for the collection of tribute by the Russian Governor. Voykar is an exceptional settlement because of its location on the crossroads between the Russian West and the Ural region and North, East and South Siberia. The main migratory flows passed through Voykar. The first written record mentioning the settlement dates to the end of the XVI and the beginning of the XVII century (a letter of AD 1601 describes a siege of the town of Berezovo in AD 1595). Besides this letter, there are census records from the XVIII-XIX centuries. These records show that the Voykar settlement was a well-developed and flourishing town as far back as the 17th-18th Centuries, with the town only being abandoned at the beginning of 20th century. Today, the settlement is completely overgrown by grass and there are no accurate dates for its foundation or rebuilding phases.

The settlement, built on permafrost, is situated on the southern shore of Voykar lake (65°41'N, 64°41'E) on the northern taiga (Fig. 1). In 2003 the Yamal Archaeological Expedition carried out an excavation at this place At the beginning of the excavation, the site looked like an 11-m high by 200-300 m diameter overgrown grass hill surrounded by a stand of spruce (Picea obovata Ledeb) established in the middle of the XIX century. Surrounding this hill is a mixed forest of larch (Larix sibirica Ledeb.) and pine (Pinus sibirica Du Tour).

Figure 1: Research area

The Expedition identified a large number of construction timbers from which wood samples were taken for tree-ring analysis. The study's aims were as follows:
- To develop three tree-ring chronologies based on living trees, growing around of the site;
- To determine the species of collected archaeological samples;
- To establish calendar dates for the settlement's main buildings;
- To develop a long tree-ring chronology by overlapping tree rings from historical timbers
with living trees.

Materials and methods

Living trees

We collected 18 cores from living spruce (termed the ELM chronology) and 12 cores from living larch (termed the LMY chronology) during the summers of 1997 and 2000. The trees grow on the southern side of Voykar Lake mixed with wild rosemary, sphagnum and pleurotium within 5-10 km of the Voykar site.

Archaeological wood

About 250 cross-sections were collected from 14 Voykar constructions and loose logs during two seasons of settlement excavation. From each construction not less than 5 wood samples were taken. Most of the samples were taken from fences, wattle fence posts and planking. Walls of buildings have few logs, because houses had only 1-3 or 1-3 joists per floor. The timbers were all well preserved in the permafrost. Bark rings were present in most samples
allowing precise identification of felling dates. All constructions were located at different heights on different parts of the Voykar hill.


We identified the tree species used in constructions at the Voykar settlement based on well-known anatomical features of spruce, larch, pine and birch (Benkova and Schweingruber 2004).

Numbers of tree rings in each cross section varied from 30 to 240 and it was possible to cross-date most of the samples found during the two years of our excavations. Samples were 'polished' using a blade and measured along two radii to reveal partial or false rings. Samples were dated with TSAP software (Rinn 1996) and the COFECHA program provided quality control for the calendar dates supplied by our new chronologies: generalised master chronologies were constructed using the ARSTAN negative exponential curve method (Holmesetal. 1986).

Samples from the Voykar settlement were crossdated with our living-tree chronologies. The spruce and larch master chronologies, 300 and 450-years-long respectively, were based on trees growing within a 5 km radius of the excavated site. In addition, earlier samples were cross-dated with а 1000-year-long larch master chronology constructed from buried wood from the South Yamal peninsula by Hantemirov (1995).


Tree species of archaeological timberwood samples

Micro-anatomical examinations of the wood show that most of the ancient samples were spruce and larch, with spruce being the most common species in this archaeological monument (Fig. 2).

Figure 2: Ratio of tree species used for construction of the Ust-Voykar settlement

A total of 111 spruce samples were collected. Spruce was used to provide fences, banks, wicker-fence posts and floors. Spruce logs were found to be particularly abundant around constructions. Unfortunately, these samples usually possessed few rings (<50) and many knots so they could not be cross-dated with any confidence.

Fifty-seven larch samples made up one third of all the samples collected over the two excavation seasons. Most of the larch samples (41) were taken from constructions, with this species used for main walls and floors in houses in the Ust-Voykar settlement. Sometimes larch logs were found mixed with spruce logs in fences.

A few floorboards in Voykar buildings were of Siberian pine, but pine wood was badly preserved. Unfortunately, we could not cross date even well preserved floor samples with 90-150 tree rings since the pine ring series showed very low sensitivity. It is a well-known fact that Siberian pine growing in the high latitudes of the Polar Urals is hard to cross-date (Shiyatov 1986). Consequently, individual pine chronologies are lacking from the dating result in this article.

Plenty of birch posts were found around later constructions in the top strata of our excavation site. However, birch was not used for main constructions i.e. houses and only a few posts were found in a wicker-fence. Unfortunately, the birch posts only had from 4 to 20 tree rings, so it was impossible to obtain tree-felling dates and, consequently, birch could not be used in further analysis. Thus, neither the pine nor the birch samples could be used and only the most numerous species, spruce and larch, were selected for further analysis. Most settlement constructions were either of larch or spruce. Samples of other species such as birch or pine were seldom employed; although a mixture of species was present: for example, in wicker-fences and in-fillings requiring regular renovation.

Statistical characteristics of tree ring series

Numbers of tree rings in our archaeological wood samples varied from 30 to 150, with an average of about 90 rings (Tab. 1). The modern larch tree-ring widths, within 3 km of the site, are narrower than those found in the larch used for construction in the archaeological monument. A comparison of the coefficient of sensitivity between the archaeological and living trees shows that the average value of sensitivity of the ring series constructed from the archaeological samples is low compared with the sensitivity of living larch growing around Voykar Lake. Comparison of the median and maximum values of the coefficient of sensitivity shows that living larch trees are characterised by high sensitivity, because these two values are closely related. By contrast, there is a big difference between the median and maximum values of the ring series from the archaeological samples (the median value is 1.7 times less than the maximum one). However, mean and median values are equal meaning that half the samples are characterised by low sensitivity (below 0.28). This suggests that the larch trees used in the constructions were probably not growing in the vicinity of the settlement, but were brought from other areas characterised by more favourable conditions for larch growth. By contrast, spruce rings from the archaeological timbers share very similar statistical characteristics, both as regards tree-ring width and series sensitivity, with the rings of living trees. The sensitivity of most spruce samples both from archaeological and living trees is low, with this being typical for the area. (Shiyatov 1986). This means that people used local spruce trees for building the settlement. Although most of the samples of spruce and larch had enough rings for the construction of time series and show typical averages for this area and adequate coefficients of sensitivity for cross dating, the sensitivity of almost half the larch samples and most of the spruce samples is low; with this consequently reducing confidence in cross dating.

Table 1: The statistical characteristics of individual chronologies of samples.
Med is median value. PAC1 is pooled auto correlation of the 1st order.

Ident Total time span Yrs Mean years Tree-ring width, mm Sensitivity PAC1
Mean Med Min Max Mean Med Min Max
LMY 1538 1999 462 267,2 0,39 0,32 0 2,93 0,42 0,41 0,31 0,54 0,58
ELM 1717 1996 280 167,8 0,76 0,72 0,05 4,34 0,27 0,27 0,17 0,43 0,58
Larch 1111 1886 810 88,15 0,63 0,58 0 3,4 0,29 0,28 0,17 0,48 0,72
Spruce 1171 1887 716 87,5 0,61 0,56 0,06 3,4 0,25 0,24 0,14 0,43 0,72
Cross dating of archaeological samples

Calendar dates for tree ages were established with the help of TSAP and COFECHA software. Ninety four out of 111 spruce samples were successfully cross-dated, with a dating success of 82 % (it should be noted that 74 of the samples were from different constructions). The dating success of the larch samples reached 90 %: with the life span of 57 larch samples limited to 50 years: 41 of these samples were from constructions.

Figure 3: Life span of trees

The results of dating are shown in figure 3. Mass tree felling for settlement construction was carried out at the beginning of the XV century (flooring and fence posts located at the bottom of the research hill were cut in 1412); throughout the XVII and half the XVIII centuries (1600 to 1730). One construction framework was found at ground level 10 m from the archaeological monument for which the trees were felled in 1850. In addition, logs dated of the second half of the XIX century are trees with the latest felling dates being 1886 and 1887 (spruce and larch respectively). But these logs do not belong to any constructions of the settlement. Trees from the top strata of the excavated site were felled in the XX century. It was possible to develop and extend the spruce and larch tree-ring chronologies using the archaeological wood (fig. 4). The larch chronology covers 810 years, and the spruce chronology 716 years extending from the 12th to the 19th centuries. Comparisons of these chronologies with living tree chronologies show good agreement.

The constructed tree-ring chronologies were overlapped with the chronologies from the living trees growing around of the Voykar settlement. The larch chronology, including living trees, covers 888 years with the archaeological samples prolonging the chronology by 427 years. The age range of the spruce chronology, including both living and archaeological samples, is 825 years. Thus, the spruce chronology is extended by 546 years.

Figure 4: Averaged spruce and larch tree-ring chronologies

Micro anatomical analysis of the most recent ring structures shows that trees used in the construction of this archaeological monument were felled in different seasons. All the wood used for large constructions such as house walls had completely-formed latewood rings under their bark. Thus, the wood for the main walls of the Ust-Voykar settlement was cut during autumn or winter, between August and May.

Timber for flooring, fences, and wicker-fence posts was cut at the beginning of summer. These samples have a few earlywood cells in their last tree ring and latewood cells are absent.

The generalised characteristics of all 14 constructions at this site are combined in Table.2. Most of the excavated constructions were built in the XVII century, a period according to the chronicles during which Voykar was intensively expanded and developed. New constructions were built every 20-30 years, with intensive building continuing until 1730. In future excavations it is hoped that we will find even more and earlier constructions.

Table 2: Calendar dates of Voykar settlement constructions.
C. - number of construction, L - Larix, P -Picea, B- Betula, S - summer (June-July), W- winter (August-May).

  Name of construction Tree species Felling year Season of tree felling
1 Fence
2 Flooring
S, W
3 Filling of C.7
4 Wall C.5
5 Bank C.6
6 Filling of C.4
7 Wicker- fence C.4
8 Framework C.7
9 Framework C.6
10 Fence
11 Filling of C.6
12 Filling of C.8
L, P, В
13 Flooring
14 Framework on the ground


1.Four species including spruce, larch, Siberian pine and birch were used to build the Ust-Voykar settlement, with locally grown spruce being the most commonly used while larch was used more sparingly.

2. Chronologies were constructed for spruce and the larch, 825 and 880-years-long respectively. There were insufficient samples of Siberian pine and birch with enough tree rings for the development of chronologies..

3. The settlement was established at least 200 earlier than is recorded in the chronicles. The oldest construction was found at the bottom of the hill. The settlement was regularly rebuilt every 20-40 years from 1412 to 1880.


The author wishes to thank Irina Panyushkina and Vanessa Winchester for their kind remarks and good advice on this article. This work was supported by the Russian Fund of Basic research (RFBR) No 05-06-80-233.


  • Baillie, M.G.L. (1982): Tree-ring Dating and Archaeology. The University of Chicago press. Chicago. 265 p.
  • Benkova, V.E., Schweingruber, F. H. (2004): Anatomy of Russian woods (an atlas for the identification of trees, shrubs, dwarf and woody lianas from Russia). Bern, Stuttgart, Wien. Haupt. 456 p.
  • Brusnicyna, A.G. (2003): Gorodische Ust-Voykarskoe. Nachalo izucheniya. Ugry. Materialy VI Sibirskogo simposiuma " Kulturnoe nasledie narodow Zapadnoy Sibiri". Tobolsk: P. 45-52.
  • Hantemirov, R.M. (1995): A 2,305 year tree-ring reconstruction of mean June-July temperature deviation in the Yamal Peninsula // International Conference on Past, present and Future Climate (Proceedings of the SILMU conference held in Helsinki, Finland, 22-25 August 1995). Ed. by Henkinheimo Pirkko. Publication of the Academy of Finland. Painatuskeskus 6/95.
  • Holmes, R.L., Adams, R.K., Fritts, H.C. (1986): Tree-ring chronologies of western North America: California, eastern Oregon and northern Great Basin, with procedures used in the chronology development work, including users manuals for computer programs COFECHA and ARSTAN. Chronology Series VI. Laboratory of Tree-Ring Research, University of Arizona, Tucson.
  • Kolchin, B.A. (1963): Dendrochronology of Novgorod. Novye metody v archeologii. Trudy Novgorodskoy archeologicheskoy ekspedicii. M. AN SSSR. No117. T3. 331 p.
  • Rinn, F. (1996): TSAP - time Series Analysis and Precipitation, Version 3 reference Manual. Heidelberg.
  • Shiyatov, S.G. (1986): Dendrochronologiya verchney granicy lesa na Urale. M. Nauka: 136 P.
  • Shiyatov, S.G., Hantemirov, R.M. (2000): Dendrochronologicheskaya datirovka drevesiny kustarnikiv iz arheologicheskogo poseleniya Yarte IV na poluostrove Yamal: drevnosty Yamala. Ekaterinburg-Salehard. Vypusk I.
© Ямальская археологическая экспедиция, 2003-2017
Яндекс цитирования