Identication of Climatic and Soil Biocenotic Factors Inuencing the Height Growth of Lapland Pine in North European Russia

Background: Lapland pine ( Pinus sylvestris ssp . lapponica Fr. ex Hartm) is the geographical and climatic ecotype and subspecies of Scots pine. It is widespread in the north of Eurasia. Height growth is interconnected with both climatic parameters and the state of the habitat of pine trees. Methods: Long-term data on height growth indices of Lapland pine (var. nana Pallas (1784)), growing in various humid biogeocenoses of three specially protected natural territories of the North European part of Russia were studied. Also sixteen basic climatic parameters averaged over the growth period of the examined trees were calculated for these regions. The comparison of the values of both different climatic parameters and the height growth of pine stands of various biogeocenoses was made by the method of cluster analysis. Results: It was established that climatic parameters such as the mean daily average temperature in January and the amount of precipitation in the spring and early summer periods had a primary influence on the cluster similarity of the height growth of Lapland pine in the North European Russia. The proximity of soil and biocenotic conditions also influenced the similarity of height growth indices of Lapland pine, but had a lower rank within the two main clusters, distinguished by climatic values. Conclusion: Our studies showed that it is possible to identify the rank influence of the most significant climatic factors and soil-biocenotic conditions on the height growth of the geographical ecotype Lapland pine subspecies with the cluster analysis.


Background
Forest ecosystems play an important role in the life of our planet. They "provide ecological, economic, social and aesthetic services to natural systems and humankind, including refuges for biodiversity, provision of food, medicinal, and forest products, regulation of the hydrologic cycle, protection of soil resources, recreational uses, spiritual needs, and aesthetic values. Additionally, forests influence climate through exchanges of energy, water, carbon dioxide, and other chemical species with the atmosphere" (Bonan 2008). Wood species are closely related both to the climate of the regions in which they grow and to the soil that nourishes and saturates them with moisture through the root system (Rysin and Savelyeva 2008). In turn, the soil supporting the life of trees itself is formed under the influence of various climatic parameters, which are among the main soilforming factors. The vital activity of trees, therefore, depends both on macroecological conditions, which are determined by climatic factors of a vast geographical area, and on microecological -soilbiocenotic factors formed at a particular point in geographical space. The annual growth rate of pine trees is a function of weather conditions (temperature and humidity) both of the current year and several previous years. This dependence is especially evident in extreme conditions of pine growth, including sites at the northern border of its range (Elagin 1976;Rysin and Savelieva 2008).
The species Pinus sylvestris L. has a high degree of polymorphism, i.e. the presence of a wide variety of intraspecific forms, which are distinguished by the most diverse signs (Pravdin 1964).
The largest taxonomic ranks within this species (subspecies) are distinguished on the basis of the geographical area of individual population growth. The main forest-forming species in the middle and most of the northern taiga of the North European territory of Russia is Lapland pine (Pinus sylvestris ssp. lapponica Fr. ex Hartm), one of the geographic subspecies of Pinus sylvestris L. (Pravdin 1964;GD 2020). Subspecies Lapland pine is widespread in Europe and Asia north of 61-62° N. This is a low plant with a maximum height of 20 m. There is also a wide variety of morphogenesis determined by the soil ecological conditions of growth, which already have lower taxonomic ranks and stand out at the species level of Pinus sylvestris L. within the main subspecies.
One of the main types of pine response to various environmental conditions is its linear growth (=height growth). Its variability is closely related to climatic factors and ecological conditions (Koukhta 2003(Koukhta , 2009Chernogaeva and Kuhta 2018;Jansons et al. 2013a, b;Pozdnyakova et al. 2019;Zhou et al. 2019). The height growth of Scots pine in cold and moist regions has been limited by temperature in the previous summer and length of the growing period (McCarroll et al. 2003;Pensa et al. 2005;Salminen and Jalkanen 2005). Whereas height growth of pine in southern regions are restricted by amount of precipitation and available water, showing positive correlation with summer precipitation and negative correlation with summer temperature (Dobbertin et al. 2010;Mutke et al. 2003;Thabeet et al. 2009). Despite the importance of the linear growth indicators in studying various stands and their changes, interest to it among research communities is much lower than to the radial growth of trees (Jansons et al. 2013a;Sánchez-Salguero et al. 2015;van der Maaten et al. 2017;Misi et al. 2019). Identifying the effects of each of these factors is a difficult but important task for understanding the relationships in the "climate-soil-plant" system in various biogeocenoses. Most often, correlation and regression analyzes are used to identify the relationship between environmental factors and linear growth. Cluster analysis occupies a special place in multivariate statistical analysis, but it is practically not used in assessing the influence of ecological factors on plant trait changes. However, the advantage of the method is that it allows to compare qualitative and quantitative features and to build their classification systems (Gitis 2003).
The aim of this work is to assess the soil biocenotic and climatic characteristics of three different areas of Lapland pine growth, as well as to identify the relationships between individual soil and climatic parameters and the linear growth of this subspecies in the North European part of Russia with cluster analysis.

Studied area description
Studies were carried out in three specially protected natural territories (SPNTs) located in the North European part of Russia: the Kivach state nature reserve (SNR) (KNR 2020), the Polar Circle state nature complex reserve of regional significance (SNCR) (BCNKC 2020), and the Pechora-Ilych state nature biosphere reserve (SNBR) (PISNBR 2020) (Fig. 1). Accoding to climatic classification of B.P. Alisov (1956), the Polar Circle SNCR and the Kivach SNR are part of the northwestern subregion of the Atlantic-Arctic forest region of the temperate zone, and the Pechora-Ilych SNBR is located in the northeastern subregion of the Atlantic-Arctic forest region of the temperate zone, which differs from the northwestern by a greater continental climate (Fig. 1, Alisov 1956). The basis of the climatic zoning and allocation of climatic regions are the features of the radiation regime and atmospheric circulation. The climate of investigated regions is characterized as cold and quite humid.
According to the soil-geographical zoning, all studied SPNTs belong to the European-West-Siberian taiga-forest region of podzolic and sod-podzolic soils of the Boreal belt (Dobrovol'skii and Urusevskaya 2004). By floristic zoning, all the studied protected areas belong to the Holarctic kingdom, the Boreal kingdom, and are included in the Circumboreal or Euro-Siberian-Canadian floristic land areas (Voronov et al. 2002).  Alisov (1956) and specified accoding to data by N.A. Myachkova (MCZR 2020) The Kivach SNR is located on the northwestern coast of Onega Lake in the southeastern region of the Baltic (Fennoscandinavian) crystalline shield. The territory of the reserve is characterized by a complex relief, formed as a result of tectonic and glacial activities. Ridge-hilly relief forms interspersed with lacustrine-glacial plains and swampy depressions (Fedorets et al. 2006). The variety of landforms determines the complexity of the soil cover. The territory is included in the Karelian province of alpha-humus podzols and bog soils of the subzone of podzolic soils of the middle taiga (Dobrovol'skii and Urusevskaya 2004). The most biocenoses of Kivach reserve are represented by different pine forests (Fedorets et al. 2006). The territory is located on the border of the ranges of subspecies of Pinus sylvestris L.: P. sylvestris ssp. sylvestris L. and P. sylvestris ssp.

Stand data
We investigated the annual variability of the indexed linear growth of Lapland pine growing in humid ecotopes, which were distinguished according to the classical typology of V.N. Sukachev (Sukachev 1972). Types of pine biocenoses were determined based on the work (Rysin and Savelieva 2008). The determination of soils for each examined area was carried out using the following works (Egorov et al. 1977;Degteva and Lapteva 2013;Fedorets et al. 2006

Climatic data
For all three areas of the study we also calculated 16 climatic parameters that could affect the linear where T a is the annual sum of active temperatures (SAT), T n is the average daily air temperature above the temperature threshold T 0 which in this work is 10°C for a time period d corresponding to the number of days in a year (Popova et al. 2017); The number of days in a year with average daily temperature above 5°C (NDY>5°C), days; The number of days in a year with average daily temperature above 10°C (NDY>10°C), days; Hydrothermal coefficient (HTC) of Selyaninov: where r n is the daily precipitation sum in the growing season (PGS) of the calendar year n; ΣT n is the sum of the average daily active air temperatures with a threshold of 10ºC for the same period of the year (May-September) (Selyaninov 1928); Simplified aridity index (SAI) of Budyko (Sirotenko and Pavlova 2012): where ∑T >10° is the annual sum of active air temperatures above 10°C (SAT>10°C); r I-XII is the annual amount of precipitation (TAP).

Statistical analyses
Statistical processing was performed using the OpenOffice.org Calc table processor and the statistical computing environment R. Generalized data for the entire observation period for individual sites with increased moisture content ("humid biotopes" according to the Sukachev classification (1972)) in the three studied regions were compared using hierarchical cluster analysis with R function hclust (method "complete")) (Koukhta 2011).
Subsequently, based on the calculated climatic parameters for various protected areas, using the hierarchical cluster analysis, the similarity of their climatic conditions was evaluated. Cluster hierarchical analysis and dendrogram construction were performed using the SciPy package for the Python 3 programming language. The analysis was performed using the complete method of calculation of the distance between clusters (Raschka and Mirjalili 2017). The Euclidean distance was used as the metric. The Euclidean distance (d) was calculated accoding Equation 4: where p and q are the measured variables (i.e. values of climatic parameter of two compared plots); i indicates the individual measurement value of the variable of a total number of n measurements.
When clustering on several grounds, they were scaled using the min-max normalization method (Raschka and Mirjalili 2017).

Results and discussion
Cluster analysis of annual average values of linear growth indexes series variability of bog ecotype of Lapland pine in the humid habitats of the three examined SPNTs for the entire observation period revealed a separation of these indicators into two main similarity clusters (Fig. 3) Thus, we see that the climatic conditions, in general, had a more significant effect on the clustering of the selected signs of linear growth. Within these two main clusters, there is a similarity between the linear growth indices of the Lapland pine in test plots with similar soil and biocenotic conditions. So, the clusters of similarity of bog podzolic soils, raised peat bog soils, transitional peat bog soils and humus peat bog soils, illuvial ferruginous gleyed podzols are clearly distinguishable (Fig. 3). This indicates that the influence of the ecological niche, especially its edaphic component, also affects the linear growth of pine, but has a lower rank relative to the climatic effect.
The next stage of our research was the identification of those climatic parameters that influenced the separation of the values of the average annual variability of the indexed series of linear growth of the stands of Lapland pine into two different clusters. The calculated values of the selected climatic parameters averaged for the period 1991-2010 are given in Table 1. This period covers both the time of ground-based measurements of the linear growth of Lapland pine and the period of growth of this subspecies before the start of measurements, but subsequently taken into account when conducting field observations. After calculating the climatic parameters presented in table 1, we also performed a cluster analysis of similarities of the three studied SPNTs using both all values of climatic parameters and each separately (Fig. 4-6).  The results of clustering of all climatic parameters in general showed that the most similar across the entire set of values were the Kivach SNR and the Pechora-Ilych SNBR (Fig. 4). The same data were obtained for the much number of other climatic parameters (Fig 4 and 5).
However, when conducting a cluster analysis by the climatic parameter mean annual air temperature (MAAT), the Polar Circle SNCR and the Pechora-Ilych SNBR turned out to be the closest (Fig. 6). This indicator brings the more northern (Polar Circle SNCR) and more continental (Pechora-Ilych SNBR) regions closer together.

Conclusion
Our studies showed that cluster analysis can be used to identify the rank influence of climatic and soil-biocenotic conditions on the height growth of the Lapland pine subspecies. Also, with its help, it is possible to reveal the most significant climatic factors for a given geographical ecotype.
According to our results, the most significant impact on the annual variability of the height growth of Lapland pine was caused by precipitation in the spring and early summer, as well as the average daily mean temperature in January. A significant effect of the average January temperature on the height growth indicators of Lapland pine was revealed by us for the first time.  Figure 1 Geographical location of study territories: A -Kivach SNR, B -Polar Circle SNCR, C -Pechora-Ilych SNBR.

Figures
Climatic zones: II -Subarctic zone, III -Temperate zone. Climatic regions: 5 -Atlantic, 8 -Atlantic-Arctic, 9 -Atlantic-Continental European, 10 -Continental West-Siberian. Borders of climatic zones and regions marked accoding to the classi cation of B.P. Alisov (1956) and speci ed accoding to data by N.A. Myachkova (MCZR 2020) Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors. Cluster analysis of the values of climatic parameters of the three SPNTs studied, averaged for the period 1991-2010 (All climatic parameters; TAP -total annual precipitation, mm; PJlS -precipitation in July-September, mm; PCS -precipitation in cold season, mm; PGS -precipitation in the growing season, mm; MTM -mean temperature in May, °C; MTWM -mean temperature in the warmest month (July), °C).
Studying sites: A -Kivach SNR, B -Polar Circle SNCR, C -Pechora-Ilych SNBR. The abscissa indicates the Euclidean distance Figure 5 Cluster analysis of the values of climatic parameters of the three SPNTs studied, averaged for the period 1991-2010 (SAT>5°C -Sum of active temperature above 5°C, °C×days; SAT>10°C -Sum of active temperature above 10°C, °C×days; NDY>5°C -the number of days in a year with temperatures above 5°C, days; NDY>10°C -the number of days in a year with temperatures above 10°C, days; HTC -Selyaninov's Hydrothermal coe cient; SAI -Simpli ed aridity index). Studying sites: A -Kivach SNR, B -Polar Circle SNCR, C -Pechora-Ilych SNBR. The abscissa indicates the Euclidean distance Figure 6 Cluster analysis of the values of climatic parameters of the three SPNTs studied, averaged for the period 1991-2010 (MAAT -mean annual air temperature, °C; MTCM -mean temperature in the coldest month (January), °C; PSp -precipitation in spring, mm; PAJn -precipitation in April-June, mm). Studying sites: A -Kivach SNR, B -Polar Circle SNCR, C -Pechora-Ilych SNBR. The abscissa indicates the Euclidean distance