UDC 571.53. 902.674_630*55

V. I. Voronin

Siberian Institute of Plant Physiology and Biochemistry SB RAS

132 Lermontov St., Irkutsk, 664033, Russia

E-mail: bioin@sifibr.irk.ru

During archaeological research in September 2007, under the foundation of the bell tower of the Spasskaya Church, the first stone structure in Irkutsk, a lezhnya made of larch logs of good preservation was uncovered. One of them had a podkorovoe ring, which made it possible to determine the year of cutting this tree. For the purpose of dating, a reference tree-ring chronology was created for 12 high-aged larch trees from the vicinity of Irkutsk with a length of 514 years (1493-2007). It was established that trees for lezhnya were harvested in the period from autumn 1752 to summer 1753. The obtained centuries-old tree-ring chronology for larch can be used to date historical wooden objects Irkutsk.

Keywords: Spasskaya tserkva, Irkutsk, lezhni logs, dendrochronology, dating.

Introduction

When determining the exact dates of construction of wooden historical objects, the method of dendrochronology is practically non-alternative. This is clearly confirmed by the already classic dendrochronological dating of such objects in Novgorod, Pskov, Smolensk, etc. [Kolchin and Chernykh, 1977], and Mangazei (Shiyatov, 1980).

In September 2007, under the foundation of the bell tower of the Saviour's Church at a depth of approx. 2 m was opened lezhnya foundation of larch logs of good preservation. The logs were located in sandy, moist soil of alluvial origin. Above them was a layer of broken bricks mixed with lime. The excavation exposed the end parts of four decking logs lying on a crossbar. The average diameter of the logs was 80 cm. One of them (N 1) still has a podkorovoe ring, which allowed us to determine the year of felling the tree. It is known that in 1672 a wooden church was built in the Irkutsk prison in the name of the icon of the Saviour Not Made with Hands, and in 1706 a stone church of the same name was laid [Levi et al., 2003]. Initially, the church was smaller in volume: the bell tower topped with a golden spire was added later, as was the stone two-story aisle on the north side. The discovered lezhnya was most likely located under the foundation of the bell tower, which was built in the late 50s-early 60s of the XVIII century. To verify this assumption, dendrochronological studies of lezhni log cuts were carried out.

Research material and methodology

With the help of an electric saw, all the end parts of the logs were cut down. Fresh cuts of two of them had a natural color characteristic of live larch wood (Fig. 1). The wood of the selected samples had a very high humidity.

page 78

Fig. 1. Ends of logs of lezhni after heating of wood samples.

To avoid destruction of the samples during drying, they were thoroughly soaked with PVA glue and then covered with a cloth for more uniform drying at room temperature. This treatment made it possible to bring them to an air-dry state without significant structural changes. The relative values of all annual layers are preserved when the samples are dried (Kolchin and Chernykh, 1977).

Principles of dendrochronological dating

The main method for determining the calendar age of trees is cross-dating. Woody plants growing in a particular region react equally to changes in external factors and have similar patterns in fluctuations in the annual growth rate. Therefore, the majority of trees within a region that is homogeneous in climatic conditions show a synchronous change in the width of their annual rings. The size of such areas can be quite large. In the Far North, where the growth of trees is determined by the temperature of the growing season, a synchronous change in the width of annual rings is observed at a distance of up to 600-800 km (Vaganov et al., 1996). In the southern regions of the forest zone, where precipitation is the limiting factor for the growth of woody vegetation, the size of such areas is significantly smaller - up to 100-300 km (Andreev et al., 1999).

When cross-dating, narrow rings are particularly indicative, indicating that the growth rate was most limited by one or another external factor. The alternation of narrow and wide rings in time is unique, so it is possible to combine the graphs of fluctuations in annual growth in the compared trees only within a strictly defined section of the dendrochronological scale. Cross-dating involves comparing graphs and determining the exact location where they match. This makes it possible to make relative and absolute dating of tree rings in the compared trees. Relative dating allows you to identify rings that were formed in the same year, and calculate how many years earlier or later one tree was cut down compared to another. With absolute dating, the calendar date of the formation of a particular ring and, accordingly, the cutting of this tree is determined. The amount of overlap between the datable and reference chronologies depends on the sensitivity and synchronicity of the samples. For reliable dating, it is necessary to overlap one curve with another on a segment of at least 50 rings (Kolchin and Chernykh, 1977).

In addition to cross-dating, dendrochronology uses the technique of comparing the sequence of reference years (reference year rings). In such years, annual rings are formed with minimal growth, which can most often be associated with unfavorable climatic conditions or with significant damage to the crown by needle-eating insects [Ibid.; Schweingruber, 1993]. Each one - or two-year growth inhibition alternates with periods of normal growth in a special way, resulting in a specific pattern characteristic of trees in one climatic region of growth. The coincidence of the reference annual rings of two growth graphs in the area of 80-100 years is a reliable estimate of complete synchronization.

Methodology for measuring and statistical data processing

The width of annual rings was measured on a semi-automatic LINTAB installation (manufactured in Germany) equipped with a computer program for processing dendrochronological data TSAP [Rinn, 1996], which allowed cross-dating of the obtained tree-ring chronologies (DCS). Annual rings were measured at four to five radii. For each of them, separate DCS were constructed, which were cross-dated to exclude cases of" falling out " of annual rings, since there were areas with rotten wood on the surface of the cut, where the structure was not clearly visible. After cross-dating of housing and communal services

page 79
for all radii, the data of measuring the width of annual rings were averaged, and an individual chronology of the tree as a whole was created, which does not have an exact time reference ("floating").

For absolute dating of archaeological logs, it was necessary to obtain tree-ring chronologies of high-aged living trees with a known date of formation of the last annual ring. Wood for the foundation of the bell tower of the Spasskaya Church, located in the center of Irkutsk, was harvested, apparently, in the immediate vicinity of the construction site. Therefore, living trees for dating should also be found in the vicinity of Irkutsk. In October 2007, drilling samples of wood*were collected from 12 high-growth larches growing within a 10 km radius of the city center. After measuring the width of the annual rings and cross-dating individual housing and communal services of these trees, a generalized absolutely dated chronology was created with a length of 514 years (1493-2007), which made it possible to date the archaeological logs, since, according to historical data, the trees for the base of the bell tower were cut down approximately 250-260 years ago.

When creating tree-ring chronologies, it is necessary to take into account the influence of age on radial growth and exclude it. Annual ring width measurements were standardized using the DPL98 software package for dendrochronological data analysis (Holmes, 1998), after which the ring width was expressed in increment indices. The indexed data have the same average values of the relative values of the annual growth variability and approximately the same variability within individual calendar intervals (Shiyatov, 1986; Metody..., 2000).

Results and discussion

Since not all archaeological logs had crustal rings, it was initially necessary to make a relative dating. This procedure was initially performed visually using the TSAP program, and then the results were checked by the COFECHA program from the DPL software package [Holmes, 1998]. The verification showed a satisfactory relative dating of the archaeological trees (Fig. 2).

On average, the correlation coefficient between the dated DCS exceeded 60 %. After their standardization, a generalized "floating" chronology was created for five archaeological trees, which can already be dated absolutely, with the definition of the calendar year of formation of the preserved subcrustal annual ring, using the DCS obtained from living trees as a reference.

The absolute dating procedure consisted in determining the matching sections of the graphs of the reference and dated chronologies. For this purpose, such techniques as visual comparison, correlation analysis, and dating by reference years are used (Kolchin and Chernykh, 1977). We took advantage of all of them. Visual comparison and search for the greatest correlation of graphs is carried out using a computer (TSAP program), as well as identification of reference years (COFECHA program). Since the approximate time range is known - the middle of the XVIII century, when the bell tower of the Savior Church was being built-the task was largely simplified. At the same time, it could not be ruled out that the archaeological logs could belong to the wall of the Irkutsk prison, laid in 1661, since the building of the Saviour's Church itself was built into it. During the dating process, this version was not confirmed.

The best correspondence of the growth graphs was found when comparing the most recent annual ring of the reference chronology dated from 1752 (tree root ring N 1) with the reference chronology dated from 1752 (Fig. 3). The overlap zone was 153 annual rings (1599-1752), which is quite sufficient for reliable dating. The correlation coefficient for this entire time interval is 0.3 (statistically significant at P < 0.01), and in the area of the greatest coincidence of the growth graphs (1673-1724; 51 year ring) - 0.58 (statistically significant at P < 0.01). We did not expect a greater correlation. First, the generalized DCS of archaeological trees is obtained from a critical number of samples (five) and is not free from" noise " introduced by individual trees, while the generalized DCS of living larches is more representative (12 trees). Secondly, the archaeological trees were most likely cut down in one small area of forest, and their DCS reflected local external influences. Live larches were selected over a larger area, in areas with different edaphic and hydrological conditions, and also affected to varying degrees by fires over the past centuries. All this could not but reduce the level of communication between the studied DCS.

To assess the similarity of the dated and reference chronologies and clarify cross-dating

* The author expresses his gratitude to E. E. Bryukhanov, an employee of Irkutsk State Technical University, for his professional assistance in selecting wood samples.

page 80

Figure 2. Fragments of tree-ring chronologies of five archaeological trees (the time period in which all chronologies are present).

Figure 3. Fragments of the reference (black) and dated (white) tree-ring chronologies in the overlap area. Arrows show a number of reference years.

their synchronicity was also determined. The more often the trees compared simultaneously form either narrow or wide rings, the higher the synchronicity between them. It should be borne in mind that the sequence of annual rings on different trunk cuts and even different radii of the same cut very often does not completely coincide. Therefore, when synchronizing, we are not dealing with the search for identity, but with establishing the degree of similarity, which can only approach 100 % [Ibid.]. The synchronicity coefficient was calculated according to the well-known Huber formula [Ibid.] and was equal to 70% for the compared chronologies, which is sufficient to confirm high-quality dating.

Additional confidence in the correctness of dating was given by a good match of the sequence of reference annual rings (some of them are shown in Figure 3). These were the narrowest rings, whose width deviations from the average values were greater than 2σ.

Conclusion

Thus, we have established the date of formation of the subcrustal ring of the archaeological tree No. 1-1752. It is likely that all the logs found in the excavation were prepared in 1752-

page 81
1753. Wooden foundations and wall connections are structures that could only have been constructed during the construction of the building itself [Kolchin, Chernykh. 1977]. However, the established date of cutting down trees for making lezhni is not necessarily the date of construction of the bell tower of the Saviour's Church. Dendrochronological dating determines the year when the last outer ring was formed, after which the tree was cut down in the period of time until the next growing season (in our case, for example, from the autumn of 1752 to the summer of 1753). The forest for the construction of other structures was freshly cut, but sometimes it was allowed to stand for one or two years [Ibid.].

Also of practical value is the generalized DCS obtained for larch trees with a length of 514 years (1493-2007). You can use it to date historical wooden objects in Irkutsk.

List of literature

Andreev, S. G., Vaganov, E. A., Naurzbaev, M. M., and Tulokhonov, A. K., Registration by annual pine rings of long-term fluctuations in atmospheric precipitation, Selenga runoff, and Lake Baikal level, Dokl. RAS, 1999, vol. 368, No. 3, pp. 400-403.

Vaganov E. A., Shiyatov S. G., Mazepa V. S. Dendroclimatic studies in the Ural-Siberian Subarctic. Novosibirsk: Nauka Publ., 1996, 246 p. (in Russian)

Kolchin B. A., Chernykh N. B. Dendrochronology of Eastern Europe, Moscow: Nauka Publ., 1977, 129 p.

Levy K. G., Zadonina N. V., Berdnikova N. E., Voronin V. I., Glyzin A.V., Yazev S. A., Baasanjav B., Ninzhbadgar S., Balzhin A. B., Buddo V. I. Modern geodynamics and heliogeodynamics. Irkutsk: Publishing House of Irkutsk State Technical University. un-ta, 2003. - Kn. 2: 500-year chronology of anomalous phenomena in nature and society in Siberia and Mongolia. - 383 p.

Methods of dendrochronology: textbook.-method. manual / S. G. Shiyatov, E. A. Vaganov, A.V. Kirdyanov, V. B. Kruglov, V. S. Mazepa, M. M. Naurzbayev, R. M. Khantemirov. Krasnoyarsk: Krasnoyarsk State University, 2000. 1: Fundamentals of dendrochronology. Collecting and receiving tree-ring information. - 80 s.

Shiyatov S. G. Datirovka derevennykh sooruzheniy Mangazei dendrochronologicheskim metodom [Dating of wooden structures of Mangazei by dendrochronological method]. Mangazeya: Mangazeysky morskoy khod / ed. by M. I. Belov, O. V. Ovsyannikov, V. F. Starkov. - L.: Gidrometeoizdat, 1980. - ch. 1.- pp. 93-107.

Shiyatov S. G. Dendrochronology of the upper forest boundary in the Urals, Moscow: Nauka Publ., 1986, 136 p.

Holmes R.L. Dendrochronology program library - users manual. -Tucson: Laboratory of Tree-Ring Research, Univ. of Arizona, 1998. - 130 p.

Rinn F. TSAP Version 3.0. Reference manual. Computer program for time series analysis and presentation. - Heidelberg: Frank Rinn Distribution, 1996. - 264 p.

Schweingruber F.H. Jahrringe und Umwelt-Dendrookologie. - Birmensdorf: Eidgenossische Forschungsanstalt fur Wald, Schnee und Landschaft, 1993. - 474 s.

The article was submitted to the Editorial Board on 27.11.08.

page 82

Permanent link to this publication: