Using collected samples 9 site chronologies of Scots pine and Siberian spruce radial increments were constructed for 5 sites. The expressed population signal (EPS) of chronologies was calculated for the common interval (1954 - 2003). Five of nine chronologies (Tab. 3) are within the well-accepted EPS threshold of 0.85 (Cook & Kairiukstis 1990). However some of the chronologies are covering longer time interval, starting from 1786, and this time interval cannot be used for the extraction of common climatic signal due to the fact that prior 1954 EPS statistics were worse. Therefore the only 5 site chronologies (Fig. 2), that are representative of the true population during the common period, were selected for the dendroclimatic analysis.
Responses to mean monthly temperatures and monthly precipitations sums
In western part of Middle taiga zone the statistically significant positive response of Scots pine to the monthly mean temperatures of November year prior to growth and May of the current growth season was found (Fig. 3B). There was no other significant response to temperature (Fig. 3A, C, D, E).
Response function analysis showed positive (Fig. 3A, D) and negative (Fig. 3B, C) response of the radial increment to the monthly sums of precipitation. Radial increment of Siberian spruce was influenced by November precipitations of the year prior to growth in eastern part of Middle taiga zone (Tab. 2, site 4) and June precipitations in eastern part of Middle taiga zone. Radial increment of Scots pine was influenced by September precipitations of the year prior to growth in eastern part of Middle taiga zone and June precipitations in eastern part of Middle taiga zone (Tab. 2, site 4). In western part of Middle taiga zone (Tab. 2, site 3) the response of Scots pine and Siberian spruce to the precipitations was opposite (Fig. 3 B, D).The increase in precipitation positively affected the radial increment of Spruce and negatively the radial increment of Pine. The similar difference between Pine and Spruce in response to the monthly sums of the autumn precipitations of the year prior to growth was found in eastern part of Middle taiga zone (Fig. 3A, C). The difference in response to the monthly sums of precipitations within the Middle taiga zone is explained by the geographical position of the sites (Fig. 1).
No statistically significant responses to temperatures and precipitations were found for Scots pine in Southern taiga zone (Fig. 3E). The lengths of growing season and annual evapotranspiration (Tab. 1) is longer here, comparing with Middle taiga zone. This is also confirmed by the smallest amount of variance explained by climate (Tab. 4).
The variance explained by the mean monthly temperature varied from 26% to 48% (Tab. 4). The highest explained variance by temperature is in most northern site chronology of Scots pine, the lowest in Southern taiga. The variance explained by the monthly sums of precipitations varied from 26% to 44%. The variation of radial increment explained by the precipitations is highest for both species in eastern part of the Middle taiga zone. There is a clear gradient of increasing amount of variance explained by monthly precipitation sums from south to north (Tab. 4). The total amount of variance in radial increment explained by climate varied from 43% to 70%.
Temporal stability of growth-climate relationships
The variables identified in dendroclimatic analysis as significant for the radial increment of Scots pine and Siberian spruce in Middle taiga zone were retested for the temporal stability of the relationships using 25-year window correlation. In this step we were looking to disprove (p>0.05) a correlation between a monthly climate parameters and ring-width chronologies for a given 25-year period if r
The statistically significant changes in correlations between climate variables and radial increment were identified for Siberian spruce (Fig. 4). In eastern part of the Middle taiga zone the correlation between precipitations of November year prior growth changed since 1987 and was significantly non stable years after 1987. The relations to temperature of June also changed in the western part of Middle taiga zone in 1987. The dendroclimatic relationships of Scots pine were also changing over the time, but those changes were statistically not significant.
The amount of variance of radial increment explained by the monthly mean temperatures and monthly precipitation sums changed over the period 25 years from 14% to 20% (Tab. 5). In western part for Siberian spruce and eastern part of Middle taiga zone for Scots pine the amount of explained variance increased on 7% and 14%. But other 3 chronologies showed decrease in explained amount of variance from 2% to 20%. Therefore using those ring-width chronologies and monthly climate data for the period from 1954 to 2003 it is not possible to make conclusion on changes in dependency of the radial increment from climate.
Climate change analysis
Fig. 5 shows the deviation in degrees C from the long-term mean (70 years) temperature and in millimetres from the mean annual sum of precipitation at the 3 meteorological stations included here with the longest period of observation. There is clear evidence during recent decades that the climate in the Komi Republic has changed. During the last 20 years the mean annual air temperature has increased, and over the last 40 years the annual precipitation sum at meteorological stations has also increased. But the impact of each month on the annual result differs over time (Fig. 6).
In some months the trends are positive, and others have some negative trends. As shown above for the boreal zone, the radial growth of trees could be attributed to certain months. Deviations from means in temperature, which are significant for radial increment of pine and spruce in Komi, are shown on Fig. 6. In our study the significance of those climatic parameters for radial increment was identified by response function analysis. The decrease in temperature of May in western part of the Middle taiga zone most probably was compensated by increase of temperature in November of the year prior to growth and increase in precipitations of June (Fig. 6). The positive response of Siberian spruce in the western part of Middle taiga zone to the precipitations of June was reflected in tree ring width (Fig. 5, Fig. 6).