Changes in foliage’s biomass of two-needled pine subgenus (Pinus spp.) and genus Betula spp. along the gradients of winter temperature and precipitation: inter-genera paradox in the forests of Eurasia


 Background

The main pool of publications on this topic is related to the assessment of possible changes in vegetation growth under the influence of climate, but few of them actually took account the impacts of global change on species composition and morphological (taxational) structure, so led to an unanswered question, how the biological productivity of the forests will change if air temperature and/or precipitation change up to a certain extent. This is a subject of the study.
Methods

In this study, our database is used in a number of 2,110 sample plots for pine and 510 for birch. In each sample plot, the biomass of the forest stands was positioned in maps of January mean temperature isolines and to mean annual precipitation ones, and the input data matrix was compiled in which the values of biomass components and of stand taxation characteristics are mated with corresponding values of climate indices. The matrix was then subjected to regression analysis.
Results

It is stated, in cold and insufficiently moisture–rich climate zones, temperature increase causes a decrease in biomass of Pinus foliage, and in other regions its increase, but the Betula pattern is the opposite. With an increase in precipitation, the Pinus foliage biomass in warm zones increase, and in cold ones it decreases, but the Betula pattern is the opposite.
Conclusion

The biomass of pine and birch stands change in gradients of winter temperature and precipitation as propeller-formed but opposite patterns, which can be explained by the different winter physiology of evergreen and deciduous species.


Introduction
In recent decades, numerous world-wide studies have proved that the irreversible climate change would result in biota catastrophe and a problem with the survival of mankind (Emanuel et al., 1985, Toman et al., 1996, Behrensmeyer, 2006 , and how these altered forests will affect the gas composition of the atmosphere changed due to climate shifts. In other words, we are dealing with a typical rigid feedback scheme that is common to all complex biological systems (Abramova, 1968, P. 169).
Long before Charles Darwin, Russian academician Ivan Lepekhin proposed the idea of changing the properties of plants and animals under the in uence of environmental changes. In the descriptions of his expedition across the Ural Mountains in Russia, he noted: "Plants and animals can get used to different climates and, depending on the habitat conditions, they can acquire properties that make their behavior reborn" (Lepekhin, 1780. P. 93). Though there is the common agreement that vegetation responses to climate change are species-speci c (Schulze & Mooney, 1994, Spathelf et al., 2018, how the biological production of stands for certain tree species (genera) would be impacted in the Trans-Eurasian climatic gradients of temperature and precipitation is unknown, because the available regional information is fragmentary and contradictory It was reported that in Canada the temperature of January and humidity conditions had a positive effect on the growth of Betula papyrifera, and the growth of Picea mariana was positively affected by the winter and spring temperatures in the current year (Huang et al. 2010). High temperature and precipitation in growth season could depict positive impact on tree biomass (Zhou et al. 2008). Furthermore, extra sugar is stored in winter tree, which will be available for tree growth till next summer (LaMarche 1974). The dendroclimatological analyses showed that the positive moisture balance in current April and previous September was the dominant climatic factor favoring the radial growth of Quercus cerris in Albania (Stafasani, and Toromani, 2015). Some discrepancies in the regional responses of the studied species mean that regional patterns of stand productivity may be not necessarily extrapolated to the transcontinental level (Huang et al. 2010).
The dependence of the radial growth of European spruce, Scots pine and European beech trees on the climate and water balance of the soil was studied using data from 24 sample plots in Germany for the period from 1951 to 2006. By combining the obtained models with climate forecast data published by the IPCC expert group for the period up to 2100, a forecast of the dynamics of growth of three tree species up to 2100 is made. It turned out that for spruce, environmental conditions become more and more unfavorable over time, which leads to a gradual decrease in growth. For Scots pine and for beech, the negative effect on radial growth of simulated climate scenarios and soil water balance could not be detected until 2100 (Röhle et al. 2010). All of the above uncertain and contradictory patterns indicate that the results obtained were not reproducible. The lack of reproducibility of scienti c results is a critical feature of modern science (Ioannidis 2005). A special study found that more than 70% of researchers tried and failed to reproduce the experiments of another scientist, and more than half of them failed to reproduce their own experiments (Baker 2016). At the same time, the limits of reproducibility in science are discussed (Guttinger 2020).
A paradoxical conclusion was reached in Russian Siberia with respect to forest cover (Lapenis et al., 2005), where with a warming climate and a simultaneous decrease in precipitation, the share of assimilation mass decreases, and the share of wood components increases. Even more, it was found that in pure forest stands, net primary production reacts to temperature increases in different climatic zones in different ways: it increases in temperate forests, remains stable in boreal forests, and decreases in wet forests in the Mediterranean, but as the biodiversity index increases, different zonal trends gradually transform into a common, uni ed negative trend. These patterns do not yet nd any biological or ecological explanation (Paquette et al., 2018). In comparison with these uncertainties and contradictions, we obtained a much more surprising and paradoxical result when studying changes in the biomass of pine and birch foliage in the hydrothermal gradients of Eurasia, which is described in this article.
The two-needled subgenus (Pinus spp.) (the diploxylon, or hard pines) is divided into the predominantly Eurasian and Mediterranean section Pinus, composed of subsections Pinus and Pinaster, includes about 100 species spread in boreal and mid-latitude zones and also in the mountain regions of the subtropical zone of the northern Hemisphere (Gernandt et al., 2005). There are about 10 species in Russia. Of the two-needled pines, the Scots pine (Pinus sylvestris L.) is the most common in Eurasia and among conifers only larches occupy a bigger area than pines. This is a large evergreen whorlbranching light-demanding tree with a transparent crown. Its needle foliage is adapted to conservative water consumption, tolerates temperatures of -50 °C to + 50 °C and lives for 5-6 years long. At the Northern limit of Eurasia, its growth is limited by a small crop of cones and very low seed quality (Boychenko, 1970). The bark is thick, scaly dark grey-brown on the lower trunk, and thin, aky and orange on the upper trunk and branches (Bobrov, 1978, Mamaev, 1983. Birch belongs to Betulaceae family. This family has 120-150 species (Grimm and Renner 2013). Majority of species are present in northern zones and have a wide natural distribution area on the Eurasian continent, ranging from the Atlantic to eastern Siberia. Genus Betula spp. is among ten common species in Russia, and 40 its species are presented in Russia. It plays an important ecological role in the formation of woody vegetation throughout the Quaternary period (Denisov, 2002), and in dry steppe conditions with high ground water forms pure stands that are resistant to droughts (Perov, 2008). There are several species in the common birch category from the section Albae Rgl To reduce the uncertainty results mentioned above, a transcontinental level of analysis was chosen in our study. The purpose of this study was to show how much the foliage biomass of the light-coniferous subgenus Pinus spp. can change with a possible increase in temperature by 1 °C at constant precipitation and with a possible increase in precipitation by 100 mm per year at constant temperature, and to compare the results with similar data for small-leaved genus Betula spp. in Eurasia.
Since the productive and carbon-deposing potential of a plant community is determined by the biomass of assimilating organs, we have focused in this article not on the total biomass of forest stands, but only on the mass of foliage.

Material And Methods
The database used in this study involves 3,020 sample plots with Pinus and 650 sample plots with Betula stand biomass in tons per hectares, including both pure stands and stands with admixture of other species (Usoltsev 2020). Since the response to climate change differs between pure and mixed stands (Paquette et al., 2018), in our comparative study we used harvest data only from pure stands that were selected from our database in a number of 2,110 sample plots for pine and 510 for birch. Subgenus Pinus sp. is presented in 86% by Scots pine (Pinus sylvestris L.) and in smaller quantities by the following species: P. tabuliformis Carr. P. densi ora S. et Z., P. nigra Arn., P. pinaster Aiton, P. brutia var. pityusa (Steven) Silba, and P. thunbergii Parl. Of the 650 sample plots for Betula, about 80% are presented by white birches, i.e. silver birch (Betula pendula Roth) and downy birch (Betula pubescens Ehrh.) We are not able to distinguish in our database these two types of birch forests, differing in their ecology and possibly in their response to climate change (Hynynen et al. 2010). However, they are very adaptive to changes in their growth environment (Laitakari 1935, Köstler et al. 1968, Ostonen et al. 2007).
In most cases, sample trees were taken on each of sample plots in a number from 5 to 10 copies. Then samples were taken from each biomass component to determine the dry matter content (and for wood and bark of stems also to determine the basic density) after drying the samples at the temperature of 80-100 0 C. The quantity of each biomass component per 1 ha was determined by regression method. Nevertheless, some sampling procedures for estimating biomass of tree components differed between the studies, since they were performed by representatives of different scienti c elds in forestry. The locations of sampling sites are shown in Fig. 1. We can see that despite the different number of sample plots for Pinus and Betula, the coverage of the territory of Eurasia by these genera is approximately the same.
In each sample plot, the biomass of the forest stands was positioned in accordance to January mean temperature isolines and to mean annual precipitation ones, and the input data matrix was compiled in which the values of biomass components and of stand taxation characteristics are mated with corresponding values of mean January temperature and precipitation taken from World Weather Maps (2007). The matrix was then subjected to regression analysis.
The question may arise why modelling was performed at a level of genera, and not for individual pine and birch species. If we adhere to the concept of species-speci c responses of forest biomass to changes in the main climatic characteristics, then when we reach the transcontinental level, we are faced with the obvious fact that no species grows throughout the continent, precisely because of regional climate differences. Moving from refuges under the in uence of geological processes and climate changes, the particular species adapted to changing environmental conditions, forming a series of vicariate species within a genus (Hultén 1937, Tolmachev 1962, Chernyshev 1974). This gives grounds for analyzing the response of tree species to changes in climate characteristics, to combine them into one climate-dependent set within the entire genus, since differences in ecological and physiological properties of different species of the genus, for example, Betula pendula Roth vs. B. utilis D. Don vs. B. maximowicziana Regel are derived from regional climatic features.
It is well known that when estimating stem biomass growth by using the annual ring width, the greatest contribution to explaining its variability being made by summer temperature. Moreover, it was established that this relationship is positive with the maximum intra-annual temperature and negative with the minimum and average annual temperature ( In terms of regression analysis, a weak temporal trend of summer temperatures compared to a steep trend of winter ones means a smaller regression slope and a worse ratio of residual variance to the total variance explained by this regression. Obviously, taking the mean winter temperature as one of the independent variables, we get a more reliable dependence having the higher predictive ability. To ensure the maximum stability of the model, each of the selected factors (independent variables) should be presented in the maximum range of its variation (Usoltsev 2003). In our example, mean January temperatures ranged from − 40 °C in the forest-tundra of the North-Eastern Siberia to + 10 °C in the subtropics of China. We compiled precipitation data ranging from 190 mm in permafrost regions of North-Eastern Siberia to 1,140 mm in South of China and to 2,500 in Greenland.
The matrix was used as a source of data (their fragment one can see in Table 1) in the subsequent regression analysis. It is well known that the biomass of a stand represented by a particular tree species is primarily determined by its age and morphological (taxation) structure, i.e., a set of characteristics such as age, mean height, mean diameter at breast height, the basal area, and the volume stock, which are interrelated. The problem of multi co-linearity arises in empirical modeling of biomass. One of the solutions to the problem is to harmonize the system by constructing recursive (recurrent, related) equations for the different mensuration indices and foliage biomass, in which the dependent variable of the previous equation is included as one of the independent variables of the subsequent one (Draper and Smith 1966). This approach as one of the methods for model harmonization (Jacobs and Cunia 1980), provides a multivariate conditionality of factors that provide exibility and universality of the regression system describing the dynamics of biomass of stands. * Designations here and further: A = stand age, yrs; V = stem volume, m 3 /ha; N = tree density, 1000/ha; P f = foliage biomass, t per ha; Tm = mean January temperature, °C; PRm = mean annual precipitation, mm.
As for choosing the structure of the regression model, we adhered to the concept that there is only one de nite variant of stand biomass structure corresponding to a given structure of taxonomic parameters (morphological structure) of a tree stand (Usoltsev 2007). The initial structure of the model included the main mass-forming indices of stands -age, stem volume, tree density, mean diameter and mean height. Mean diameter, and mean height were excluded in the process of the regression analysis as these factors were not statistically signi cant. The synergism (lnA) · (lnN) was introduced in the model to account for the decrease in the tree density with age and its effect on the foliage biomass. The nal structure of the model included only those mass-forming indices that were statistically signi cant for foliage biomass component.

Results
The    Table 3. We can see that mass-forming variables explain in averaged about 78, and 61% of the variability of foliage biomass in Pinus, and Betula correspondingly, including 47, and 39% of the contribution from the stem volume. Climate variables explain about 22, and 39% of the foliage biomass variability in Pinus, and Betula correspondingly, i.e. less than about 3.5, and 1.6 times as much as mass-forming variables in Pinus, and Betula correspondingly.  (Fig. 2).

Discussion
We can see in Fig. 2 that there are two patterns of signi cantly differential changes in the pine and birch foliage biomass in precipitation and winter temperature gradients in the form of two oppositely directed propellers. The most interesting question is how much the foliage biomass would change with an air temperature deviation from the usual norm, for example, by 1 °C and with a precipitation deviation from the usual norm, or by 100 mm per year. The constructed models give us the answer to such question in relation to pine and birch stands. To do this, we took the rst derivative of 3-Dimensional surfaces (Fig. 2), and not analytically, but graphically, i.e. we took off the biomass difference interval (Δ, %%) corresponding to temperature interval 1 ℃ and precipitation interval 100 mm directly from the graphs or from the corresponding tables, and get the answer in the form of 3-Dimensional surfaces divided into plus and minus areas that correspond to the increase or decrease in the biomass of stands having the age of 100 years for Pinus (Fig. 2a) and 50 years for Betula (Fig. 2b)   We can see in Fig. 2a, that pine forests show a monotonous increase in foliage biomass in the gradient of temperature increase but only when there is su cient water supply (PR m = 900 mm), and as the transition from moist areas (PR m = 900 mm) to areas of nonsu cient water supply (PR m = 300 mm), the trend changes to the opposite (Fig. 2a). In the gradient of increasing precipitation, pine forests show a monotonous increase in foliage biomass but only in cold areas (Tm = -30 °C), and as the transition from cold areas to warm ones (Tm = 10 °C), the trend changes to the opposite (Fig. 2a). A similar general pattern was observed earlier at the local level in the marsh forests of Siberia, where at the maximum amounts of temperature sums above 10 °C (2200 °C) there is an increase in the radial growth of stems by 30-50% with an increase in precipitation from 400 to 600 mm, and at the minimum amounts of temperature sums (1600 °C) the radial growth is reduced by 4-9% with an increase in precipitation in the same range. Correspondingly, at the level of precipitation of 400 mm the radial growth is reduced by 14-20% with an increase in the sum of temperatures from 1600 to 2200 °C, and it increases by 14-33% in the same temperature range at the level of precipitation of 600 mm (Glebov and Litvinenko, 1976). According to the results obtained by Molchanov (1976), in the North of Eurasia the greatest in uence on the growth of the annual tree ring is the air temperature, and in the conditions of the southern forest-steppe the dominant role is played by precipitation.
It is well known the Liebig's law of the minimum (1840), according to which a growth rate depends on the factor that is at the minimum in relation to its needs. Although J. Liebig, followed by J. Esslen (1905), had shown that a limiting factor can be not only a lack, but also an excess of such factors as light, heat and moisture (a lot of "good" is also "not good"), nevertheless, he focused his attention mainly on the effect of the minimum of chemicals (oxygen, phosphorus, boron, etc.), and as a result of that, this phenomenon was established in science as the law (principle) of the minimum by Liebig. The idea of the limiting in uence of the maximum on a par with the minimum was developed by V. Shelford (1913), who extended the limiting principle to any environmental factors and became known as the author of Shelford's law of tolerance. W.P. Taylor (1934) followed the same concept. Later A.A. Molchanov (1971) interpreted the limiting principle in relation to forest ecosystems as an "extended concept of limiting factors", according to which "any state approaching or exceeding the limit of resistance for any organism and groups of interest can be considered as a limiting factor" (p. 271). Recently, this phenomenon has be-come widespread as the principle of limiting factors by Liebig-Shelford (Rozenberg et al., 2016). The reaction of pine foliage biomass (Fig. 2)  consequently, the limits of applicability of the models for pines and birches are the same. However, the biomass of birch stands in the database is 4 times less than that of pine, but even if we equalize the harvest data pools for pines and birches, the limits of applicability of the models will not change, and, of course, this will not lead to a change of signs in birches for independent variables -temperatures, precipitation, and their synergy. Then what can be the reason for the opposite trends in foliage biomass in pine and birch, which are statistically signi cant probability level reaches 0.9999, i.e. p-value < 0.00001.
Let us try to link the obtained counterintuitive patterns of changes in foliage biomass with previously published Trans-Eurasian trends in foliage productivity as the ratio of annual growth in the stem volume or aboveground biomass to the biomass of the assimilation apparatus. It is known as foliage e ciency (FE) (Usoltsev et al., 2018). It was found that in the direction from the northern moderate zonal belt to the subequatorial one, FE increases in pine and decreases in birch in the same zonal range. This phenomenon seems to be related to the fundamental difference between the winter physiology of evergreen and deciduous species, namely the ability of the former to assimilate atmospheric carbon dioxide and prolong the assimilation process beyond the vegetation period, which is usual for deciduous species (Jumelle, 1892, Ewart, 1896, Matthaei, 1902, Henrici, 1921, Iwanoff, Kossowitsch, 1929, Zacharowa, 1929, Ivanov & Orlova 1931, Printz, 1933, Cartellieri, 1935, Ålvik, 1939, Freeland, 1944, Zeller, 1951, Pisek, Rehner, 1958, Pisek, 1960, Lyr et al. 1974, Kramer & Kozlovsky 1983, Schaberg et al., 1995, Wieser, 1997, Smashevskiy, 2014. In pine trees in the southern direction, due to higher winter temperatures, the winter accumulation of assimilates increases as a result of excess of photosynthesis over respiration, which is associated with the autumn-winter litter fall of foliage (Zalesov, et al., 1994), which apparently determines the trend of increasing FE in the direction from the northern temperate zone to the subequatorial one.
In birch trees, FE in the direction from the northern temperate to the subequatorial zone does not increase, as in pine, but decreases, possibly due to an increase in the respiration losses during a shorter physiologically active period (due to leaf shedding) compared to pine. In essence, everything depends on the ratio of photosynthesis and respiration in the pine, especially during the period when the birch does not have photosynthesis, but only respiration. If we assume the presence of a positive ratio of foliage biomass and FE, then in our case, the opposite FE trends of two species coincide with opposite trends in foliage biomass, but only in regions of su cient moisture (PRm = 800-900 mm) (Fig. 2a, b), and as we move to areas of insu cient moisture (PR m = 300 mm), this coincidence is replaced by a complete contradiction.
The differences between pine and birch in FE values are also apparent as the climate continentality (summer aridity of the climate) increases: FE decreases most intensively in pine forests, and signi cantly less intensively in birch communities. As the continentality index by Khromov (1957) increases from 55 to 95%, the FE in pine decreases by 8 times, and in birch the corresponding decrease in FE is less than 2%. The decrease in foliage biomass coincides with the decrease in FE as the climate increases in aridity only in pine trees and only in warm regions (T m = 0 to10 °C), and when moving to cold regions, this coincidence is replaced by a complete contradiction. For birch, the decrease in leaf biomass as the climate increases in aridity coincides with a decrease in FE only in cold regions (T m = -30 to -40 °C), and as the transition to warm regions, this coincidence is replaced by a complete contradiction.
The patterns of biomass amount change under assumed changed climatic conditions (Figs. 2-4) are hypothetical. They re ect long-term adaptive responses of forest stands to regional climatic conditions and do not take into account rapid trends of current environmental changes, which place serious constraints on the ability of forests to adapt to new climatic conditions The main pool of our harvest data on forest biomass in Eurasia were obtained since 1970s to 1990s, and the climate maps used cover the period of the late 1990s and early 2000s. Some discrepancy between the two time periods may cause possible biases in the results obtained, but for such a small time difference in the initial data, the inclusion of compensatory mechanisms or phenological shifts in forest communities is unlikely (Anderegg et al., 2019, DeLeo et al., 2020). There is an uncertainty in assessing the impact of phenology on the biological productivity of stands, established for the cherry oak in the South of Russia: if the assessment of the biomass of oak stands did not reveal differences between the phenological varieties of oak, then the assessment of net primary production shows a 1.6-fold advantage of late-blooming variety over the early-blooming variety (Zhou Wen Nan, 1992).
Taking into account the stated methodological and conceptual uncertainties, the results presented in this study should be considered as preliminary ones. They can be modi ed if the biomass database will be enlarged by additional site-speci c and stand-speci c data. A full explanation for both the obtained regularities and counterintuitive results can be obtained after conducting detailed physiological studies.

Conclusions
A comparison of the reaction of pine and birch foliage biomass to changes in the average January temperature by 1 °C at constant precipitation and average precipitation by 100 mm per year at constant temperature showed counterintuitive result. If low temperature and heavy precipitation are the limiting factors for pine leaf biomass, then low temperature and insu cient precipitation are the limiting factors for birch. When the temperature increases by 1 °C, the change in leaf biomass in pine, which is negative in dry conditions, becomes positive in wet conditions, and in birch, a positive change in dry conditions becomes negative in wet conditions. When precipitation increases by 100 mm, the change in leaf biomass in pine trees that is negative in cold regions becomes positive in warm regions, and in birch trees, the positive change in leaf biomass in cold regions becomes negative in warm regions.
Hovewer, from the long-term perspective, climate change might bring even more drastic modi cation of winter temperature and annual sum of precipitation than was considered here. Therefore, our outputs represent just an example of model sensitivity to changing climatic conditions. The development of such a model for the main forest-forming species of Eurasia allow us to predict changes in the foliage productivity of the forest cover of Eurasia in relation to climate change.
Declarations Figure 1 The distribution of 2,110 sampling sites of Pinus (left hand side) and 510 sampling sites of Betula (right hand side) on the territory of Eurasia.

Figure 2
Dependence of pine a and birch b forest biomass (Pf) of Eurasia upon the mean January temperature (Tm) and mean annual precipitation (PRm).

Figure 3
Simulated changes in pine a and birch foliage biomass due to temperature increase of 1°C based on the derived model (for the stands aged 100 years). Here the value 1 represents the plane corresponding to zero change of biomass at the expected temperature increase by 1°C; the value 2 represents the border between positive and negative changes in biomass (Δ, %) at the expected temperature increase by 1°C. Figure 4 Simulated changes in pine a and birch b foliage biomass due to the assumed precipitation increase of 100 mm (for the stands aged 50 years).