Valda Araminienė·Lucian Dinca·Iveta Varnagirytė–Kabašinskiene·Raluca Enescu·Vlad Crisan·Vidas Stakėnas

Abstract The effects of different climatic conditions on growth and chemical composition of Betula pendula in geographically different European regions,Lithuania and Romania,were compared.Birch species in the entire area have a wide natural distribution,but B.pendula is commercially more important in Lithuania than in Romania.Here we evaluated tree ring width,wood density and foliar chemical composition of mature birch trees in two European regions.Trees at the Lithuanian sites had greater radial growth and wood density with no clear changes in foliar chemistry than those at the Romanian sites.Mean wood density was 600—700 kg m−3 at Lithuanian and 350—450 kg m−3 at Romanian sites.Mean width of wood ring,earlywood and latewood for Lithuanian birch trees were several times higher than the means for Romanian birch trees.We hypothesized that the main differences in birch radial growth and wood density were due to the different climatic conditions in the studied regions.Ca,K and Mg concentrations were significantly higher and Fe and Mn were lower in the birch foliage at the Lithuanian sites compared those at the Romanian sites.Overall assessment of growth showed that silver birch cultivation is more appropriate for colder climate regions and that birch growth may change in the context of a warming climate.

Keywords Betula pendula · Ring width·Ood density ·Foliage chemistry

Introduction

Climate change is expected to increasingly affect the species composition of Central European forests (Lindner et al.2014;Wohlgemuth 2015) through significant effects on biological processes and plant life cycles and functions (Norby and Luo 2004).Physiological changes in plants,caused by climatic phenomena,change the chemical composition of foliage and litterfall,and,therefore,can affect soil nutrient cycle and ecosystem productivity.Changes in the composition of older leaves caused by climatic factors (Top and Filley 2014;Suseela et al.2015) can affect the activity of detritivory (Currano et al.2008;Couture et al.2012),which will have an impact on carbon and nutrient cycling in soils(Aerts 1997;Liu et al.2009;Suseela et al.2013).Furthermore,these processes can affect the overall productivity of the ecosystem (Grimm et al.2013).

Birch species are among the main tree species in Lithuania and throughout the northern and eastern parts of Europe and an essential component of temperate and boreal forests.In Europe,the most common birch species are silver birch (Betula pendulaRoth.) and downy birch (B.pubescensEhrh.),which is more frequent in the northern regions (Beck et al.2016).These two are important deciduous species and provide popular raw material for forest wood industries in the Baltic Sea Region,especially in Lithuania,Finland,Sweden,Latvia and Russia (Verkasalo et al.2007;Mantau 2012).Birch is mainly used for pulp,paper and paperboard,plywood,veneer and sawn goods,particle-and fiberboard,and firewood and chips for energy.Silver birch is an important pioneer species in Romania,which has traditionally been used in ethnomedicine and households and for other customs (Papp et al.2014),but generally is considered as a tree species of low value.

The inf ulence of the northern climate and light conditions on woody vegetation during the growth period is not fully understood,but longer days and lower night temperatures are known to enhance flavonoid production in plants (Jaakola and Hohtola 2010).The chemical composition of tree leaves is also altered by changing climate.Drought and higher temperatures can alter plant processes leading to changes in photosynthesis,stomatal conductance,leaf pigmentation,water potential,protein concentrations,and thus growth and yield(Benjamin and Nielsen 2006;Rennenberg et al.2006;Praba et al.2009;Anjum et al.2011).Pradhan et al.(2018) found that if rainfall is reduced by 28% from the present rainfall,diameter growth inB.pendulawould decrease up to 25%.According to Rubio-Cuadrado et al.(2018),growth ofB.pubescensis weakly related to climate variables.However,concentrations of quercetin derivatives in white birch foliage(Stark et.al.2008) and of monoterpenoids,proanthocyanidins and flavonols in juniper needles (Martz et al.2009)increased with latitude and altitude in Finland.

Here we evaluated growth variables (tree ring width and wood density) and foliar elemental composition of mature birch trees in geographically different European regions,Lithuania and Romania (the Transylvania region),which represent different climatic conditions and biogeographic regions.Lithuania is assigned to the southern part of the hemiboreal forest zone,and Romania,as the more biogeographically diverse country,is assigned to more southern biogeographic regions such as continental to alpine and steppe.We followed the presumption that different regions can be used as potential natural laboratories to determine the response of a forest to global warming (Jump et al.2009;Blois et al.2013).We hypothesized that birch growth will differ significantly in the two countries that differ climatically.

Material and methods

continental-influenced climate with warm,dry summers and fairly severe winters.The mean annual air temperature is 6.5 °C,and the mean annual precipitation is 686 mm.Winter temperatures can be about − 20 °C with northeasterly winds,and heavy snowfall are also possible.Summer daytime temperatures are 25—30 °C.The growing season lasts for 202 days in the western part of the country and 169 days in the eastern part.Total forest area is 2.1 million ha,or 33.5% of the land area (State Forest Service 2017).Coniferous stands are frequent in the western,eastern,and southern Lithuania:deciduous stands are in the middle of the country.Scots pine (Pinus sylvestrisL.),silver birch (Betula pendulaRoth) and Norway spruce (Picea abies(L.) Karst) represent the most common forest tree species in Lithuania and compose 35%,22% and 21% of the forested area,respectively(State Forest Service 2017).Birch stands cover about 20%of the total forest area and are considered as one of the most important deciduous tree species in the country.

Romania is located in the southeastern part of Central Europe on the lower course of the Danube River on the western coast of the Black Sea.The Romanian landscape is variable,proportional and symmetric.Mountains occupy 35%,hills 35% and plains 30% of the land area.The highest altitude is 2544 m above sea level in the Fagaras Mountains.The climate is temperate-continental with moderate features,characteristic for Central Europe.The southwestern part of the country receives warm air masses from sub-Mediterranean areas and the southeastern part is influenced by dry air masses from the southwestern Asia (Feurdean 2004).The mean annual temperature is 11 °C,and precipitation varies between 400 mm (in the plains area) and 1000 mm (in the mountains area).The warmest areas are in the southern Romania,with summer daytime temperatures of 25—30 °C.The northern and eastern mountainous districts of Transylvania are colder.Throughout the country,winters are cold with abundant snowfall,especially in the mountainous areas.Total forest area is approximately 6.5 million ha,or 27% of the land area (FAO 2020).Coniferous cover makes up about 30% of the forested area,pure and mixed beech stands 30%,oak species 19%,broad-leaved hardwood species 14% and broad-leaved softwood species 6%.Birches are found predominantly in the hilly and mountainous areas because it has generally low demands (Sof letea and Curtu 2007).

Sampling and analysis

Study areas

Lithuania,the largest country of the Baltic States,is situated along the coast of the Baltic Sea between 56° 27′ and 53° 53′ latitude and 20° 56′ and 26° 50′ longitude.The Lithuanian landscape is predominantly flat.The highest altitude is 294 m above sea level.The climate varies between maritime and continental;it is a typical European

Pure silver birch stands growing in plains at 95—201 m a.s.l.in Lithuania and birch stands at of 501—828 m a.s.l.in Romania were selected for the study (Fig.1).The soil at Lithuanian sites is classified as Luvisol and Cambisol at Romanian sites according to the World Reference Base for Soil Resources (IUSS Working Group WRB 2015).To provide basic information about the mineral soil,thepH and mean concentration of nitrogen (N) were obtained from forest management plots adjacent to the sample plots of this study.For the Romanian soils,the mineral soil pH in the humus-accumulative horizon (Ap) ranged from 3.7—3.8 (Barsa) to 4.9—6.0 (Zarnesti and Vad),and mean concentration of N was 0.5% − 0.9%.For Lithuanian soils,the mineral soil pH in the Ap horizon was pH 3.3—5.4,and mean concentration of N was 0.2%—0.3%.The trends in the total soil nutrient concentrations in the stands were similar within each country and mostly represented fertile or very fertile sandy loam soils,which are typical for birch stands in each country.The site conditions in two countries differed because of topography,soil formation factors,parent material,etc.Therefore,soil differences could be a limitation to evaluating responses in this study.

Fig.1 Locations of the study plots in Lithuania and Romania(1:Skriaudžiai;2:Pakuonis;3:Girionys;4:Vad;5:Zărnești;6:Bârsa)

Samples for the chemical analyses of silver birch leaves and stem increment cores were collected from six silver birch stands,including three plots from Lithuania and three plots from Romania (Fig.1).Selected plots represented typical forest types in each country,and the main characteristics of the study sites are presented in Table 1.

Table 1 Description of study sites in Romania and Lithuania

For this study,15—16 trees in Lithuania and in Romania were selected (3 plots in Lithuania,3 plots in Romania,5 − 6 birch trees in each plot).The birch foliage and stem cores were sampled in July—August 2017.In the field,stem cores were sampled from birch stem at 1.3 m above ground level (breast height) from the selected trees in both countries.In the laboratory,the stem cores were processed using the LignoStation wood analysis system (RinnTECH Inc.,Heidelberg,Germany),which utilized the dielectric wood properties.A high-resolution (50 μm) wood density profilewas acquired for each core using a high-frequency probe(Schinker et al.2003).For each tree ring,total ring width,latewood width,wood density of ring and wood density of earlywood and latewood were measured.

Leaves for chemical analyses were collected from selected trees in both countries after intensive leaf growth in July—August 2017.Typical leaves without any visible damage were sampled from 6 to 7 branches in the middle of the crown on the south side of trees growing in similar light intensity and temperature.Samples for an individual tree were then combined for analysis of concentrations of carbon (C) and N,P,K,Ca,Mg,Mn,Zn and Fe.Total N and C were determined using a CHN analyzer.The foliar concentrations of P,K,Ca,Mg,Mn,Zn and Fe were analysed using a wet mineralization method with strong acids,then the total concentration of each element was measured using atomic adsorption spectrometry.The chemical analyses were performed in the Laboratory of Pedology and Foliar Analysis at the National Institute of Research and Development in Forestry in Bucharest.

Calculations and statistical analyses

Means ± SE were calculated by combining all data from the three sites for all variables in each country.Normal distribution of the data was checked using Lilliefors and Kolmogorov—Smirnov tests,and the hypothesis of normality was rejected.For the above reasons and because of small amount of data,a nonparametric statistics has been chosen.For testing the differences between growth variables and foliar concentrations from two countries,the nonparametric Mann—Whitney U-test was used.Additionally,the coefficient of variation was calculated to assess the data homogeneity.For visualisation of the foliar concentrations variance between the two countries,graphs were generated using Excel (Microsoft,Redmond,WA,USA).Spearman rank correlation coefficient was also calculated.Statistical analyses were done using Statistica 7.0 (TIBCO Software,Palo Alto,CA,USA) and a level of significance ofp<0.05.

Results

Birch tree radial growth

The analysis of the annual increment in ring widths at the studied sites showed greater radial growth for Lithuanian birches (Fig.2).The differences between two sites were obvious for 1975—1997 with a clear decrease in mean birch wood ring width at Lithuanian sites.The difference between annual ring widths in the two countries remained during the studied period,but since 1997 it has not been as large as before 1997.The annual wood ring width remained relatively stable for Romanian sites compared to Lithuanian sites.

Fig.2 Mean widths of annual rings measured from cores at breast height from birch trees at the studied sites in Lithuania and Romania.Bars show SE

The mean wood ring width per 10-year period was greater for the Lithuanian than Romanian sites (Fig.3).Similarly,the width of both earlywood and latewood was also greater for Lithuanian sites.At Lithuanian sites,the mean ring width was 2.4 times higher for 1975—1984,1.9 times higher for 1985—1994,and 1.4—1.8 times higher for 1995—2016 than at Romanian sites.

Fig.3 Mean width of birch wood ring,earlywood and latewood measured from cores at breast height from birch trees for 10-year periods for the sites in Lithuania and Romania.Bars show SE

Birch wood density

Mean wood density was higher for Lithuanian sites(600—700 kg m−3) compared to Romanian sites(350—450 kg m−3;Fig.4).

Fig.4 Mean birch wood density measured from cores at breast height from birch trees at the sites of Lithuania and Romania

In more detail,the lowest mean density of birch wood was 295 kg m−3at the Girionys site,and the highest density was 901 kg m−3at the Pakuonis site.For the same period,the lowest mean density was 146 kg m−3at the Bârsa site,and the highest was 926 kg m−3at the Vad site in Romania.

Mean density of birch wood rings,earlywood and latewood per 10-year period at Lithuanian sites was 1.2 − 1.3 times higher for 1975—1984,1985—1994 and 1995—2004 and 1.6 times higher for 2005—2016 compared to Romanian sites(Fig.5).

Fig.5 Mean wood density of birch wood ring,earlywood and latewood given per 10-year period for sites in Lithuania and Romania.Bars show SE

Elemental composition of birch foliage

The mean concentrations of total carbon (C),main macronutrients (N,P,K,Ca and Mg) and micronutrients (Fe,Zn and Mn) in birch foliage in all Lithuanian and Romanian sites are given in Table 2.

The coefficient of variation for each element in Romania and Lithuania was highest for the concentrations of K,Mn and Zn,and the lowest for C concentration (Table 3).

Significantly (p<0.05) higher concentrations of K,Ca and Mg in birch foliage were found at the Lithuanian sites(Fig.6).No statistically significant difference was found for C concentration or macronutrients N and P.Significantly lower values were obtained for concentrations of Fe and Mn in the foliage of Lithuanian birch trees in comparison to Romanian trees.

Fig.6 Mean concentrations of carbon (C) and some macronutrients (N,P,K,Ca and Mg) and micronutrients (Fe and Mn) in the birch foliage at the Lithuanian and Romanian sites.Error bars represent standard deviation.Significance levels:** p ≤ 0.01;*** p ≤ 0.001

The measured correlations between different chemical elements in birch foliage showed different results for the Lithuanian and Romanian sites (Table 4).For Lithuanian birch trees,significant correlations (r=0.43—0.47;p=0.01)were found for C and Mn,Ca and Mn,K and Mn,K and Zn,also Zn and Fe.A strong correlation was found between K and Fe (r=0.66;p=0.01).For Romanian birch trees,very strong positive correlation (r=0.92;p=0.01) was found between N and P concentrations in birch foliage,and relatively strong correlations (r>0.80;p=0.01) were found between the following elements:N and Ca,P and Ca,also P and K.

Table 3 Coefficient of variation for each element at the Romanian and Lithuanian sites

Table 4 Correlation coefficients between chemical elements in birch foliage determined at the Lithuanian and Romanian sites

Discussion and conclusions

The average temperature for the European land area for the last decade (2002 − 2011) is 1.3 °C above the pre-industrial level.An increase in the highest winter temperature is projected for eastern and northern Europe and in summer temperature in southern Europe during the twenty-first century.In Lithuania,an increase in temperature of 0.27 °C per decade and decrease in precipitation of 48 mm per decade was determined for 2007—2014 (Araminienė et al.2019).For Romania,mean annual temperature in 2015 was by 1.96 °C higher than in 1961—1990 (Mateescu 2016).Followed by the mentioned climatic changes in the region,it is important to evaluate the possible impacts on forest trees species.The present study was designed to determine variations in radial growth,wood density and foliar chemistry of silver birch trees at two geographically different sites,representing different climatic conditions.The current study found that annual tree rings were wider for Lithuanian sites,representing cooler climatic conditions.Another important finding was that wood density,including both earlywood and latewood,was higher for Lithuanian sites.As reported in earlier studies,tree growth responses to climate vary according to geographic location,forest type,and tree species (Messaoud and Chen 2011;Liang et al.2019) and air temperature,precipitation and soil moisture (Scharnweber et al.2011;Beedlow et al.2013).The growth—climate relationship along latitudinal gradient was studied by Lyu et al.(2017)who found similarities of tree growth responses to growing season temperature between latitudinal and altitudinal gradients and differences in climate-growth relationships due to differences in moisture conditions.Lower temperatures limit or prevent growth,and reproduction and increase mortality of trees (Jobbágy and Jackson 2000;Henttonen et al.2014).For example,in climate chamber experiments,higher atmospheric temperature combined with increased CO2tended to intensify the growth of silver birch seedlings,which subsequently led to greater biomass (Araminiene et al.2018).Elevated temperature and CO2can improve plant growth,especially aboveground (Araminiene et al.2018).However,at higher temperature,tree competition and drought also limit tree growth and recruitment (Mäkinen et al.2003;Loehle et al.2016).Silver birch trees in the present study grew less along the latitudinal gradient.

At Lithuanian sites,mean wood density was 600—700 kg m−3and 350 − 450 kg m−3at Romanian sites.Others found a mean basic wood density for silver birch of 500 ± 20 kg m—3(Heräjärvi (2002) and 512 kg m—3(Heräjärvi 2004),375—590 kg m—3(Helińska-Raczkowska 1996),483 kg m—3(Bhat 1980),illustrating that birch wood is of medium density compared to other European hardwoods.These values are lower than those we obtained.Wood density can vary among different sites;e.g.,mean density increases with increasing tree age (Heräjärvi 2002;Lachowicz et al.2019).In our study,mean basic density was highest for the 70-year-old trees.From a practical forestry view,the best quality silver birch wood is found at 50—70 years of age,and the raw material depreciates with age of birch wood (Lachowicz et al.2019).Other studies conf irmed that the location of forest stands affect the basic wood density of silver birch (Lachowicz 2010,2015).Forest site type and location,and the interaction between these conditions have a statistically significant impact on mean wood density;silver birch wood from more fertile forest sites had lower wood density than the sites of lower fertility (Lachowicz et al.2019).Möttönen and Luostarinen (2006),comparing the basic density of birch wood along the trunk from plantations and from natural forest,reported that the mean wood density for the whole trunk was 454 kg m—3for trees in plantations and 507 kg m—3in natural forest.According to the latest studies on basic wood density of 22-year-old silver birch trees,seed origin(Baltic countries and Finland) has no influence on wood density (Viherä-Aarnio and Velling 2017).Mean wood density was 473—481 kg m—3for 30-year-old silver birch trees.Values for trees farther to the north were lower than those found in our study,illustrating a downward trend of tree wood density from the south to the north in Europe.

In the current study,we found statistically significant differences for Ca,K,Mg,Mn and Fe in birch foliage between the Lithuanian and Romanian sites.Ca,K and Mg concentrations were higher at the Lithuanian sites,whereas Fe and Mn concentrations were higher at the Romanian.There have been no case studies that compared differences chemical composition in silver birch foliage from different geographical locations.Theoretically,different geographic regions will differ in temperature and CO2levels.ElevatedCO2was associated with decreased concentrations of N,K,Cu,S and Fe in silver birch leaves (Oksanen et al.2005)and higher air temperatures with lower plant N concentrations (Weih and Karlsson 2002).The 2.5% N concentration in silver birch leaves in Finland (Aosaar et al.2016) corresponds to the mean N concentration found in our study,but they did not show a clear trend from the south to the north of Europe.Our results to some extent ref lect those of Araminienė et al.(2018) who also found that higher temperatures tended to increase the leaf N and C concentrations in silver birch seedlings.

Generally,N concentrations positively interact with P,K and Ca concentrations (Alexe 1991).The results of this study did not explain the clear differences between the two countries.For Lithuania,the correlation between K and Fe in birch foliage was statistically significant,while in Romania relatively high correlations were found between N and K and between Ca and P concentrations.It can therefore be assumed that there was no significant trend between the two countries,representing different climatic conditions.Our data also indicated that the decreases in N,and/or N/P and N/K toward the north suggest that decreased N concentration may limit birch growth.

Our results suggest that birch growing in the northern region tend to be limited by a combination of lower nutrient uptake,possibly because of poor soil conditions.However,the above results do not explain the better birch growth and higher wood density at the Lithuanian sites,which may thus be better explained by air temperature than by soil conditions and nutrient uptake.In warmer climatic conditions,silver birch grow faster (Araminiene et al.2018),which could lower wood density.Further studies are needed to assess these variables and potential relationships.

AcknowledgementsThe paper partly presents the findings obtained through the long-term research programme“Sustainable Forestry and Global Changes”implemented by the Lithuanian Research Centre for Agriculture and Forestry.