Ecophysiological Variability of Alnus viridis (Chaix) DC. Green Alder Leaves in the Bieszczady Mountains (Poland)

Alnus viridis (Chaix) DC., green alder, is a fast-growing shrub that grows expansively in the European mountainside. In Poland, A. viridis naturally occurs only in the Bieszczady Mountains (south-eastern part of the country), above the upper forest border. In this study, we assessed the potential of green alder to expand in post-farming areas in the Bieszczady Mountains. We investigated the effects of topographical, climatic, and edaphic characteristics of four various study sites on the physiological and morphological properties of A. viridis leaves in order to answer the question whether the growth of plants in lower positions improves their physiological condition to such an extent that it increases the species invasiveness. This is the first comprehensive ecophysiological study of this species to be carried out in this part of Europe. The photochemical efficiency of PSII, the chlorophyll content, and leaf 13C and 15N discrimination were analyzed. On the basis of leaf radiation reflection, coefficients such as reflectance indices of anthocyanins, carotenoids, flavonoids (ARI2, CRI1, FRI), photochemical index of reflection (PRI), and the water band index (WBI) were calculated. We observed favorable physiological effects in A. viridis plants growing in locations below the upper forest border compared to plants growing in higher locations. As a result, A. viridis may become an invasive species and disturb the phytocoenotic balance of plant communities of the altitudinal zones in the Polish Western Carpathians.


Introduction
Alnus viridis (Chaix) DC., known as a green alder, is a shrub belonging to the birch family (Betulaceae). A. viridis is a heliophilous species; it is also resistant to harsh mountain conditions, including low temperatures. White et al. [1] argued that A. viridis plays an important role as a pioneer species because it successfully colonizes areas after strong environmental disturbances. A. viridis is a woody boreal species with an Alpine-Central European area of diffusion, common at the tree line and at high latitudes. It is found in the mountains of Central Europe, specifically in France, Switzerland, Italy, and Germany to Ukraine, Romania, and Bulgaria. In Poland, green alder reaches the north-western border of the Carpathian range and has the character of the Eastern Carpathian and subalpine species [2]; its natural sites occur only in the Bieszczady Mountains (south-eastern part of plants in positions below the upper forest line improves their physiological condition (in relation to plants from higher localities) to such an extent that it increases the potential invasiveness of the species, which may result in disturbance of the phytocenotic balance of plant communities in post-agricultural areas. For this purpose, it was determined whether the elevation and nature of the habitat (including exposure to stressors) influenced leaf morphology, photosynthetic apparatus, stable isotope composition ( 13 C and 15 N) and the content of photosynthetic and photoprotective pigments in A. viridis leaves.

Leaf Morphology
The results of A. viridis, the appearance of the bush and leaf morphology, varied depending on the study site and locality (Supplementary Table S1, Figures S1 and S2). The highest basic morphological parameter values, i.e., surface area, length, width, and perimeter, were characterized by A. viridis leaves from the lowest localities, i.e., 568 m and 980 m a.s.l. (Łobozew and Przełęcz Wyżna, respectively) ( Table 1). However, the values of these parameters for A. viridis leaves growing at 980 m a.s.l. were significantly lower than those growing at 568 m a.s.l. The leaves of shrubs growing in a higher locality-1215 m a.s.l. (Połonina Wetlińska) and 1320 m a.s.l. (Tarnica)-showed the lowest values for all parameters, statistically different from the lower elevations (Table 1). The specific leaf weight (SLW) coefficient was significantly different depending on the locality elevation ( Table 1). The lowest SLW value was observed for A. viridis leaves at an elevation of 568 m, while the highest was at 1215 m a.s.l. Intermediate SLW values were found at 980 m, while at 1320 m the value of SLW was slightly lower than that at 1215 m a.s.l.

Chlorophyll (Chl) Content in Leaves
The total content of Chl (Chl a + b) in A. viridis leaves at the beginning of summer in the Bieszczady Mountains (July) was lowest in the lowest locality (568 m a.s.l.). At the highest localities (1320 and 1215 m), the content of Chl was similar, although significantly lower compared to that at 980 m a.s.l. (Table 2). In autumn (September) the total Chl content was significantly lower compared to July at the lowest elevations (568 m and 980 m a.s.l.) and it did not significantly change when compared to 1215 m and 1320 m a.s.l. In leaves from all localities, the total Chl as well as the Chl a content was higher in the leaves of A. viridis at the beginning of summer than in autumn (Table 2). However, the content of Chl b in A. viridis leaves was lower in July compared to September, except for the site at 980 m a.s.l. In September, the leaves of shrubs growing in lower localities had significantly less Chl b than those from the higher ones. Chl a/b ratios were significantly higher in the summer than in the autumn, except for the locality at 980 m a.s.l. In addition, the higher localities were characterized by a lower value of the ratio Chl a/b in relation to those located lower, which was particularly obvious during autumn months (Table 2). The analysis of 13 C discrimination showed significant differences between the studied shrubs ( Table 3). The largest negative δ 13 C discrimination values (−30.78‰) indicative of optimal growth conditions, were found in A. viridis growing near the village of Łobozew at the lowest locality (568 m a.s.l.). At other field sites, Przełęcz Wyżna (980 m a.s.l.), and Tarnica (1320 m a.s.l.), the differences in δ 13 C values were smaller ( Table 3). The pattern of changes in the δ 15 N discrimination value is very similar to the δ 13 C profile ( Table 3). The largest negative values were recorded at the locality in Łobozew 568 m a.s.l. (Table 3).

Analyses of Chl a Fluorescence
In September, the plants from all elevations analyzed were characterized by clearly elevated fluorescence values (FL) at steps J and I, noticeable on the transient fluorescence curve of Chl a, compared to those in July (Figure 1a). In autumn, curves for shrub leaves growing at an elevation of 980 m a.s.l. and higher had a similar shape with a marked increase in FL intensity in the J-I phase as compared to leaves from the lowest locality. The obtained JIP curves did not show the characteristic visible deflection for leaves at 568 m a.s.l. (Figure 1a-blue empty points). In the summer, differences in the so-called thermal phase (J-P) were particularly visible between the highest and lowest localities and between 980 and 1320 m a.s.l. (Figure 1a). The differential curves for the O-K, O-J, J-I and I-P phases were constructed analogously to ΔVt-as the difference between normalized FL values recorded in September and July, for each elevation (Figure 2). A detailed analysis of the O-J phase revealed the presence of the K-band (between 200 and 300 μs on Figure 2b), which was not visible on the O-P integral differential curve (Figure 1b). In turn, the O-K differential curve revealed the L-band, illustrating the efficiency of photon absorption and the use of energy in the initial phase of the light reaction of photosynthesis. The L-band was present on differential curves in A. viridis leaves from all these localities (Figure 2a). The differential curves (∆Vt) of FL Chl a kinetics in the O-P region for the localities (July and September measurements from all vegetation seasons) are shown in Figure 1b. The main differences between the summer-autumn period are visible in the J-I phase. However, the biggest differences occurred in shrubs from the locality at 1215 m a.s.l., where the J-band was higher and shifted in time (Figure 1b-green points).
The differential curves for the O-K, O-J, J-I and I-P phases were constructed analogously to ∆Vt-as the difference between normalized FL values recorded in September and July, for each elevation (Figure 2). A detailed analysis of the O-J phase revealed the presence of the K-band (between 200 and 300 µs on Figure 2b), which was not visible on the O-P integral differential curve (Figure 1b). In turn, the O-K differential curve revealed the L-band, illustrating the efficiency of photon absorption and the use of energy in the initial phase of the light reaction of photosynthesis. The L-band was present on differential curves in A. viridis leaves from all these localities (Figure 2a). The curves for individual sections were obtained by subtracting the Chl a fluorescence values recorded in July from the curves obtained in September. As a control for each field test stand, the curve obtained in July was adopted, as described in Methods.
However, in plants growing at lower elevations, the L-band reached values above 0.08 and was twice as high as in shrub leaves growing in the highest localities (1215 and 1320 m a.s.l.). A similar relationship was visible in the case of the K-band (Figure 2b). In contrast to July, during September, some disturbances in the oxidative-reduction balance were visible in plants from field sites at 980, 1215, and 1320 m a.s.l. (positive values in the J-I phase, Figure 2c). In the case of the lowest elevation, the reverse trend could be observed, as proven by the negative values ΔVtJ-I. In turn, stress-typical negative bands did not appear in the I-P region of the FL induction curve (Figure 2d).
The chosen parameters of the fluorescence kinetics of Chl a are presented as a percentage deviation from the values obtained for A. viridis plants growing at the lowest locality (blue line-control) (Figure 3a,b). A comparison of the parameters of PSII photochemical efficiency in July in the control versus plants growing at 980 m a.s.l. did not show significant differences in the functioning of the photosynthetic apparatus ( Figure 3a). However, leaves of A. viridis grown at 1215 m and 1320 m a.s.l. differed in term of the values of the Chl a fluorescence parameters when compared to the control. Among them, a significant decrease in the maximal fluorescence (Fm), variable fluorescence (Fv), efficiency of the water splitting complex on the PS II donor side (Fv/F0), and the maximum quantum yield of PSII (Fv/Fm) was found (Figure 3a). With the increase in elevation, there was also a decrease in the rate of energy uptake and electron transport (TR0/Cs and ET0/Cs parameters, respectively), as well as decrease in the PSII performance index (PI) in relation to the control.  The curves for individual sections were obtained by subtracting the Chl a fluorescence values recorded in July from the curves obtained in September. As a control for each field test stand, the curve obtained in July was adopted, as described in Methods.
However, in plants growing at lower elevations, the L-band reached values above 0.08 and was twice as high as in shrub leaves growing in the highest localities (1215 and 1320 m a.s.l.). A similar relationship was visible in the case of the K-band ( Figure 2b). In contrast to July, during September, some disturbances in the oxidative-reduction balance were visible in plants from field sites at 980, 1215, and 1320 m a.s.l. (positive values in the J-I phase, Figure 2c). In the case of the lowest elevation, the reverse trend could be observed, as proven by the negative values ∆Vt J-I . In turn, stress-typical negative bands did not appear in the I-P region of the FL induction curve (Figure 2d).
The chosen parameters of the fluorescence kinetics of Chl a are presented as a percentage deviation from the values obtained for A. viridis plants growing at the lowest locality (blue line-control) (Figure 3a,b). A comparison of the parameters of PSII photochemical efficiency in July in the control versus plants growing at 980 m a.s.l. did not show significant differences in the functioning of the photosynthetic apparatus ( Figure 3a). However, leaves of A. viridis grown at 1215 m and 1320 m a.s.l. differed in term of the values of the Chl a fluorescence parameters when compared to the control. Among them, a significant decrease in the maximal fluorescence (F m ), variable fluorescence (F v ), efficiency of the water splitting complex on the PS II donor side (F v /F 0 ), and the maximum quantum yield of PSII (F v /F m ) was found ( Figure 3a). With the increase in elevation, there was also a decrease in the rate of energy uptake and electron transport (TR0/Cs and ET0/Cs parameters, respectively), as well as decrease in the PSII performance index (PI) in relation to the control. A significant increase in the relative fluorescence value (Vj) in shrub leaves at an elevation of 1320 m a.s.l. was noticeable. The parameters values of plants growing at an elevation of 980 m a.s.l. were most similar to the control values ( Figure 3a). Plants growing at positions 980, 1215, and 1320 m a.s.l. (in contrast to A. viridis growing at 586 m a.s.l.) show F v /F m values lower than 0.83, which indicated a stress response of the photosynthetic apparatus. This differentiation was even greater for the PI parameter values, which better describes the physiological condition of the plant, than F v /F m . The PI value provides information on the number of active reaction centers per chlorophyll and initial reactions of the light phase with data on electron flux through RC [17] and is a good indicator of plant stress. growing at 586 m a.s.l.) show Fv/Fm values lower than 0.83, which indicated a stress response of the photosynthetic apparatus. This differentiation was even greater for the PI parameter values, which better describes the physiological condition of the plant, than Fv/Fm. The PI value provides information on the number of active reaction centers per chlorophyll and initial reactions of the light phase with data on electron flux through RC [17] and is a good indicator of plant stress.
In September the Chl a fluorescence parameter values of plants growing at the highest localities (1215 and 1320 m a.s.l.) changed slightly (with the exception of Vj) compared to those from July (Figure 3b). The largest difference was only seen for A. viridis leaves at an elevation of 980 m a.s.l.

Reflectance of A. viridis Leaves
The reflectance spectrum of A. viridis leaves in all localities analyzed was characterized by a similar curve shape ( Figure 4). In terms of photosynthetically active radiation (PAR), the differences between the intensity of reflectance in plants at individual elevations were small. The reflectance intensity on the yellow-green spectrum, with its peak at 550 nm, increased with the locality elevation (Figure 4a-inside). The biggest difference was visible on the orange spectrum and at 696 nm for plants growing at 1215 and 1320 m a.s.l., when compared to the lowest locality (568 m a.s.l.). For a shrub at 980 m a.s.l. there was a shift in maximum sensitivity towards longer wavelengths (with a peak at 704 nm), with simultaneously lower sensitivity values on the yellow-orange spectrum. The minimum value was consistently around 670 nm for plants from all localities (Figure 4c). In September the Chl a fluorescence parameter values of plants growing at the highest localities (1215 and 1320 m a.s.l.) changed slightly (with the exception of Vj) compared to those from July (Figure 3b). The largest difference was only seen for A. viridis leaves at an elevation of 980 m a.s.l.

Reflectance of A. viridis Leaves
The reflectance spectrum of A. viridis leaves in all localities analyzed was characterized by a similar curve shape ( Figure 4). In terms of photosynthetically active radiation (PAR), the differences between the intensity of reflectance in plants at individual elevations were small. The reflectance intensity on the yellow-green spectrum, with its peak at 550 nm, increased with the locality elevation (Figure 4a-inside). The biggest difference was visible on the orange spectrum and at 696 nm for plants growing at 1215 and 1320 m a.s.l., when compared to the lowest locality (568 m a.s.l.). For a shrub at 980 m a.s.l. there was a shift in maximum sensitivity towards longer wavelengths (with a peak at 704 nm), with simultaneously lower sensitivity values on the yellow-orange spectrum. The minimum value was consistently around 670 nm for plants from all localities (Figure 4c).  An analysis of spectral reflectance indices showed that the lowest value of the anthocyanin reflectance index (ARI2) was found in plants at 568 m a.s.l. In the remaining localities, the ARI2 values were similar, although significantly higher than the values at 568 m a.s.l. (Table 4). The lowest carotenoid reflectance index (CRI1) was recorded in plants at 1215 m a.s.l. (Połonina Wetlińska). In the remaining localities (Łobozew, Przełęcz Wyżna, and Tarnica), the values of this parameter were more or less homogeneous, although sig-  Table 4).

Discussion
The significantly lower (compared to other localities) values for the leaf morphological parameters of A. viridis shrubs that grew on the highest, unprotected peak and sub-peak localities (1215 m and 1320 m a.s.l.) are most likely the result of the environmental stresses that these plants undergo (such as strong wind, large fluctuations of temperatures, and intensive solar irradiation) ( Supplementary Figures S1 and S2). However, these values do not differ significantly from those given in the literature for 12 populations in the Polish Bieszczady Mountains and for four populations from Switzerland [18].
Low SLW values in the lower localities (568 m and 980 m a.s.l.) indicate that the large leaf area is accompanied by a low dry matter value (Table 1). To minimize the effects of stressors, plants usually develop leaves covered with a thick cuticle layer or trichomes. In the Bieszczady Mountains, lower localities are usually exposed to fewer strong wind and high temperature fluctuations, and are more likely to be shielded and shaded. Shading, for example, causes the plant to develop leaves with a large area, but thin leaf lamina [15,19]. The modifications in leaf morphology at the highest localities (at 1215 m and 1320 m a.s.l.) most likely result from various exposures to strong wind. The locality at 1215 m a.s.l. (Połonina Wetlińska) is positioned on a bare ridge, while the locality at 1320 m a.s.l. is found on a steep (about 45 • ) north-western slope below the peak of Tarnica. It seems, therefore, that exposure to stressful environmental factors related to a.s.l. locality influences the variability of the morphological parameters of A. viridis leaves.
The chlorophyll content in plant leaves and their respective proportions are controlled by many factors, both external and internal. The integrated activities of these factors determine the activity of the biosynthesis and degradation processes of Chl [20]. It was reported that in the leaves of A. viridis (the mountains of North Tyrol), the Chl content decreased with increasing elevation, with the highest content observed in July and the lowest in June and November. In the present study, no clear relationship was found between the height a.s.l. and the total Chl content. In terms of seasonal changes in the Chl content in A. viridis leaves, the Chl content in September was significantly lower compared to that in July only in the lower locations (568 m and 980 m a.s.l.). In this study, Chl content measurements in A. viridis leaves were conducted generally under lower elevations than in the Tyrolean mountains (650-1950 m a.s.l.). On the other hand, the differences in the seasonal Chl content might result from the fact that in lower positions, the radiation conditions vary more than in a peak locality. In lower localities, the leaves were thinner and had a larger surface area, which can result in a lower Chl content per unit area (Tables 1 and 2).
Because the elevation of the localities did not significantly affect the Chl a content either in summer or autumn, the total content of Chl a + b reflected changes in the Chl b content at these localities ( Table 2). In the early stages of plant growth, the leaf maximum Chl content was found to be controlled by phytochromes under optimum light conditions [21]. The red color of the light causes an increase in the total Chl content, and the quantitative composition of Chl a in relation to Chl b can change depending on the red to far-red ratio (R/FR). Red light usually favors the production of Chl b, which in turn leads to a decrease in the ratio of Chl a/b. In addition, the final value of this index under conditions of a high R/FR ratio (a lot of red) also depends on the age of the plant, and thus can vary within a large range, sometimes as much as from 1.0 to 2.2 in Chenopodium rubrum [21,22]. In fully exposed localities at an elevation of 1215 m and 1320 m a.s.l., the higher Chl b content in A. viridis leaves is probably due to the high R/FR ratio ( Table 2). In contrast, the low Chl b content in A. viridis leaves observed in the locality 980 m a.s.l. in September is most likely the result of a fluctuation in the spectral composition of light at this study site during the growing season. In addition, at this locality, A. viridis grows in the vicinity of other plants.
It is assumed that δ 13 C values for C 3 plants range from −20 to −35‰ (on average −27.5‰) [23,24]. In all localities, δ 13 C values oscillated near the average value, which indicates that these plants are able to bind CO 2 directly, with the predominant involvement of the ribulose-1,5-bisphosphate carboxylase oxygenase (RuBisCO) enzyme (Table 3). However, the significantly less negative values of δ 13 C at an elevation of 1215 m a.s.l. indicate the least favorable vegetation conditions prevailing at this locality, which is most likely related to the presence of stress factors affecting the closing of stomata, especially drought. The significantly lower WBI values recorded in this locality confirm the water deficit conditions and appearance of leaf dehydration (Table 4).
Based on research conducted by [25], the relationship between carbon isotope discrimination and the intensity of water loss due to transpiration has been estimated. Therefore, it can be assumed that the δ 13 C values obtained may result from the fact that A. viridis shrubs growing in the studied localities differ in terms of their water use efficiency.
It is well known that discrimination of the 15 N content in plants and soil occurs as a result of both biological and physicochemical processes [26,27]. Nevertheless, the processes conditioning the δ 15 N value in plants, and in the soil, are still poorly understood. In our study, high negative δ 15 N values correlated with high negative values δ 13 C ( Table 3). The least negative values of δ 15 N (−1.13‰) were found in leaves in the locality 1215 m a.s.l., which at the same time were characterized by the lowest δ 13 C (−26.84‰). A. viridis, thanks to symbiosis with F. alni in root nodules, has the ability to bind N 2 [5,6]. Initially, it was suggested that N isotope fractionation ( 15 N discrimination) occurs during N 2 binding in plant root nodules [28]. However, [29] demonstrated that during the N 2 binding process, discrimination is rather minimal. Significant isotope fractionation may take place during subsequent transport and biochemical transformations that occur inside the host plant [29][30][31]. Unkovich et al. [29] also pointed out that root-nodule bacteria, which infect a symbiotic plant, have an effect on the fractionation of bound N in plant organs [32][33][34]. This was confirmed by studies on various A. viridis tissues in the Swiss Alps [9]. It cannot be ruled out that the difficult growth conditions resulting from the nature of the habitat reduce the internal discrimination of 15 N.
The increase in FL at steps J and I on the Chl a fluorescence curve, as observed for all measurements in September, suggests a lower number of electron transporters on the PSII acceptor side than in July. The shape of the induction curve and the relatively high FL intensity in the J-I phase in plants at 980, 1215, and 1320 m a.s.l. indicates deviations from the oxidation-reduction balance of the Q A pool (Figure 1a), as reported before by [35]. The I-P phase illustrates the photosystem I (PSI) acceptor side reduction process [36]. The slightly lower FL intensity in the I-P phase in the localities at 980 and 1215 m, specifically visible in autumn, suggests a lower efficiency in the reduction of electron transporters by terminal electron acceptors such as ferredoxin, NADP + , and RuBP as indicated also by [11]. An analysis of the J-P phase shows that in July, the highest efficiency of electron transport from Q A to PSI was demonstrated by plants in the locality at 980 m a.s.l., which may indicate that the photosynthetic apparatus is more efficient and better acclimated to environmental stress (Figure 1a-red filled points) [37,38]. High FL values in the thermal phase J-P part of the JIP curve are related to the reduction of the final steps of the electron transport chain [39]. In both July and September, the shape of the curves for shrub leaves in the locality at 568 m a.s.l. suggests that of the localities analyzed, this is where the photosynthetic apparatus of A. viridis plants is in the best condition. Differences between individual habitats in the summer-autumn period (∆Vt) correspond to the degree of reduction in the secondary electron acceptor (Q B ), plastochinone (PQ), and cytochrome b 6 f [14] and are visible in the J-I phase (Figure 1b). The highest value of ∆Vt was observed in plants growing at an elevation of 1215 m a.s.l., which points to a strong reaction of A. viridis photosynthetic apparatus to the unstable ambient conditions prevailing at this locality.
The disclosure of the K and L-bands on the differential FL curves indicates disturbances during the course of the photosynthetic reaction in A. viridis leaves at the end of the growing season. In the autumn, the plants in all localities decreased the efficiency of energy transfer between the antenna complexes and the PSII reaction center, as evidenced by the appearance of the L-band on the O-K differential curve (Figure 2a). A. viridis shrubs in localities situated at 980 m a.s.l. (and below) showed greater disorder (higher L-band and K-band values) in the initial reactions of the light phase than those growing at higher altitudes. This is most likely related to the impairment of the PSII donor side, in particular, the partial oxygen-evolving-complex deactivation [40]. The lower ∆Vt value for the L and K bands in shrub leaves at higher localities is possibly the result of photosynthetic apparatus acclimation to the more difficult environmental conditions prevailing there throughout the year. Presumably, the decrease in the intensity of Chl a FL in the L-band is associated with an increase in energy transport efficiency between the light harvesting complex of PSII (LHCII) and the PSII reaction center [36]. The positive values for bands visible on the J-I differential curve in the higher localities are evidence of a balanced disorder between Q A reduction and oxidation (Figure 2c). The response of plants to unfavorable environmental conditions often results in negative values in the I-P region of the FL induction curve [41]. However, in A. viridis leaves, there were no bands characteristic of stress symptoms in spite of the slight decrease in the number of available NADP molecules in the PSI reaction center (Figure 2d). This observation indicates that the long-term effect of possible multi-stress is not reflected in the I-P phase. The close correlation between PSII function and Chl a fluorescence parameters allows a quick and precise evaluation of the photosynthesis response to environmental changes [39,42]. In September, in A. viridis shrubs growing at an elevation of 980 m a.s.l., the values of majority of the PS II photochemical efficiency coefficients (except for the time necessary to obtain maximum fluorescence (T fm )) approached those observed at sub-peak and peak localities (1215 and 1320 m a.s.l.). For healthy plants under stress-free conditions, the maximum value of the F v /F m coefficient is about 0.83 [43]. An F v /F m value less than 0.83 indicates that the plant has been exposed to stress factors affecting the functioning of PSII and is considered to be a sensitive indicator of PSII photochemical efficiency reduction [44]. Miszalski et al. [45] demonstrated that in Picea abies in the Tatra Mountains, increasing elevation decreased the F v /F m and F m values. However, in our research, the leaves of A. viridis shrubs were less differentiated in terms of the F v /F m parameter than PI. Consequently, in A. viridis leaves, PI values better reflect changes in the photochemical activity of PSII in response to changing environmental conditions (Figure 3a,b). The Chl a fluorescence parameter changes and differences in FL intensity in the J-I phase were similar to previous reports, which indicates that as the elevation increases, the level of environmental stress usually increases. This additional stress is probably associated with a drop in the average annual temperature and air humidity, higher solar radiation intensity, and a shortage of nutrients in the soil [14,38,41]. As the elevation increases, the daily amplitudes of environmental parameters increase, as does the role of wind [10].
An analysis of leaf reflection allows for a non-destructive and quick assessment of plant sensitivity to the effects of stress. An increase in reflection values in the photosynthetically active radiation (PAR) range, with its peak in the green spectrum, usually occurs under stressful conditions [46]. In the present study, we observed changes in the reflection intensity of A. viridis leaves as a function of increasing elevation a.s.l.-it was clearly greater at the highest localities. This was reflected in the peak heights found on the sensitivity plots. Despite the slight differences in the reflection intensity (Figure 4a), the sensitivity values had clear visible peaks, characteristic of a stress response (Figure 4c). The maximum values of the sensitivity were observed at the highest localities (1320 and 12,315 m a.s.l.). Bands at the same wavelengths were also observed on the sensitivity curves for other plant species subjected to drought stress, fungal infection, strong dehydration, or high-light stress [47][48][49]. For leaves of A. viridis that grow on Połonina Wetlińska (1215 m a.s.l.), an intense reflection was additionally observed in the far-red and the near infra-red (NIR) range (starting at 830 nm) compared to the other localities studied. This is probably the result of the lower hydration of leaves in this locality; this assumption was confirmed by the low WBI (Table 4).
Of the substances responsible for absorbing harmful UV radiation, flavonoids, anthocyanins, and carotenoids play the most important role [10]. Analyzing reflectance indices of light radiation from leaves in relation to the shrub study site is a valuable source of information about the changes in the content of plant pigments. The lowest amount of anthocyanins in A. viridis leaves (ARI2 index) was observed in the lowest growth locality (Table 4). In other populations analyzed, the content of anthocyanins was comparable. The carotenoid content, estimated on the basis of the CRI1 index, was lowest in A. viridis leaves on Połonina Wetlińska (1215 m a.s.l.). The highest flavonoid content was demonstrated for shrubs growing on Tarnica, at 1320 m a.s.l. The small number of flavonoids and carotenoids in the A. viridis leaves on Połonina Wetlińska (low values of the FRI and CRI1 coefficient, Table 4) is somewhat surprising, taking into account that these pigments are protective against UV radiation. An explanation for this fact may be the high anthocyanin content of these leaves, which may provide them sufficient protection against the harmful effects of radiation. The WBI values for different plant species usually range from 0.8 to 1.2 [50]. The WBI values obtained for A. viridis (about 0.9), therefore indicate relatively good hydration of plant tissues in the localities studied (Table 4). Even the lowest hydration level, which occurred in the leaves of plants growing at an elevation of 1215 a.s.l., can be considered to be within the norm. In turn, the PRI coefficient is correlated with zeaxanthin (de-epoxidation in the xanthophyll cycle) and the effectiveness of PAR utilization by plants [51]. Higher PRI values indicate better PAR utilization efficiency. Gamon et al. [52] showed that, using the PRI, it is possible to track changes in the effectiveness of light radiation use in the photosynthetic process of plants affected by various environmental factors (e.g., the availability of mineral substances). The best use of PAR by A. viridis was recorded in the lowest field position, at 568 m a.s.l., which may be the result of growth under near-optimal conditions.

Study Sites and Plant Material
The studies were carried out on green alder, Alnus viridis (Chaix) DC., leaves in the years 2014-2016. Research was conducted at the beginning and at the end of their growing season, i.e., in July and September, respectively. Four localities were chosen at different elevations: three in the Bieszczady National Park (BNP, south-eastern part of Poland) and one located outside its borders (Supplementary Table 1 Table S1). A. viridis is found there on the edge of the forest, in close proximity to other bushes and tall trees, creating compact clusters. The locality outside the BNP is also the lowest, in post-agricultural areas, near the village Łobozew (568 m a.s.l.). The shrubs are larger there, in terms of their morphology, than those growing on the Przełęcz Wyżna and on Połonina Wetlińska ( Supplementary  Figures S1 and S2).
Analyses were performed on intact leaves (directly on a shrub) as well as leaves detached and collected from A. viridis shrubs in the July and September months. Measurements and leaf collection were performed only on sunny days with a clear sky and 4-6 h after the beginning of the light period. Leaves were removed from various parts of the bush under comparable light conditions. For further study, usually 20 similar leaves per plant for one measurement were used/collected-in all localities over the course of two consecutive days. Plant material was collected and kept in a liquid nitrogen container for transport and analyses. The morphological variability of the leaves and the reflection of light radiation was analyzed in September. Leaves for 13 C and 15 N discrimination analyses were collected only in September. The September measurements made it possible to assess the course of physiological processes taking place at the height of the growing season and reflected the "physiological history" of the A. viridis leaves.

Leaf Morphology Parameters
Leaf shape was analyzed using the computer image analysis device WinDIAS_3, (GEOMOR-TECHNIK, Szczecin, Poland). Leaf parameters such as length, width, perimeter, and surface area were determined. Specific leaf weight (SLW) was calculated by dividing the dry weight of the leaf (mg) by its surface area (cm 2 ). The results presented in the paper are mean values from measurements made on 60 selected leaves from each locality (described above).

Chlorophyll Content in Leaves
The Chl content of A. viridis leaves was measured according to the method described by [53]. At each locality, two discs with an area of 1.76 cm 2 were cut from 10 selected similar leaves, by corkscrew, symmetrically on both sides of the primary nerve (midrib); these samples were placed immediately in test tubes containing 5 mL dimethyl sulfoxide (DMSO). After transfer to the laboratory, the tubes were heated for 3 h at 65 • C to extract Chl. The extract was then poured into spectrophotometric cuvettes, and the absorbance was measured using a CE2501 spectrophotometer, 2000 series (Cecil Instruments Ltd., Cambridge, UK) at λ = 665 and 648 nm.
The contents of Chl a and b were calculated according to the following formulas:

The Content of Carbon Isotope 13 C and Nitrogen 15 N
The lyophilized leaves of A. viridis, harvested at the end of the growing seasons (20 leaves from each locality), were ground to powder in an agate mortar. The stable carbon and nitrogen isotopes were analyzed on a Thermo Flash EA 1112HT elemental analyzer coupled to a Thermo Delta V Advantage mass spectrometer (Thermo Fisher Scientific, Bremen, Germany) in the continuous flow system. The samples wrapped in tin foil were burnt in an oxygen atmosphere at 1020 • C. The CO 2 and N 2 obtained from the combustion were separated on a chromatographic column and directly inserted into the spectrometer with a capillary tube. Measurements were calibrated via international standards USGS 40, USGS 41, and IAEA 600 [54]. The 13 C results are shown in δ form relative to the VPDB standard. The 15 N results were given in δ form relative to the N 2 standard.

Chl a Fluorescence Measurements
Chl a fluorescence kinetic parameters were analyzed using the Handy-PEA portable fluorimeter (Hansatech Instruments, Norfolk, UK), allowing for non-invasive measurements, according to [55]. Each time, measurements were made on the upper surface of 20 selected leaves. Before Chl a fluorescence measurements (actinic light 3 mmol (quantum) m −2 s −1 ), leaves were acclimated to darkness for 20 min using the measuring clips. Results were read and analyzed using the PEA Plus (Hansatech Instruments, Norfolk, UK) and Microsoft Excel 2010 software. The following parameters were analyzed: minimum fluorescence (F 0 ), F m , F v , F v /F m , F v /F 0 , T fm , surface area above the chlorophyll fluorescence induction curve (Area), PI, V J , TR 0 /Cs, ET 0 /Cs, and DI 0 /Cs. These data were used to perform the JIP test, taking the following fluorescence intensity measurement points: O-20 µs, J-200 µs, I-30 ms, and P-300 ms. The curves of fluorescence kinetics (∆Vt) were obtained by subtracting the normalized fluorescence values between the points O-P, measured in each field site in September from those recorded in July (accepted as a control), according to the formula: where Vt IX -relative variable fluorescence in September; Vt VII -relative variable fluorescence in July

Reflection of Radiation from the Leaves
The reflectance was measured each time on the upper side of the leaf blade of selected leaves using a miniature leaf spectrometer CI-710 (CID Bio-Science, Camas, WA, USA). The presented results are mean values from measurements taken on fully developed leaves from two growing seasons. The signal integration time was 350 ms; the signal smoothing factor (boxcar) was set to 10. Each measurement is the average of 10 scans. Based on the results, the reflectance intensity curves as a function of the wavelength of light were plotted. Reflectance coefficients were calculated automatically by the instrument on the basis of the following formulas: ARI2 = (R −1 550 − R −1 700 ) · R 800 [56]; CRI1 = (R −1 520 − R −1 700 ) · R 800 [56]; FRI = (R −1 410 − R −1 460 ) · R 800 [56]; WBI = R 970 · R 900 −1 [57]; PRI = (R 531 − R 570 ) · (R 531 + R 570 ) −1 [52].
R is the intensity of reflectance at the wavelength (in nm) given in the index. The reflectance difference and sensitivity allow for the identification of wavelength, at which the reflectance value changes most, under the influence of environmental stress [58]. The reflectance difference (RD) was calculated according to the formula given by Carter et al. [47]: where R X -reflectance intensity (%) between 400 and 1000 nm in plants growing at 980, 1215, or 1320 m a.s.l., strongly exposed to the effect of multistress; R 568 -reflectance intensity (%) between 400 and 1000 nm in plants growing at 568 m a.s.l.; this was used as the reference group. The sensitivity was calculated by dividing the RD values by the mean reflectance values (%) for plants growing at 568 m a.s.l., according to the formula: sensitivity = RD (R 568 ) −1 .

Statistical Analysis
All data are mean values from measurements made on leaves in individual localities during the growing seasons (2014-2016). The significance of the differences between the averages was examined using one-way ANOVA. The significance of differences between the means was compared using the Tukey test for different N at a level of p ≤ 0.05. The results were statistically analyzed and the Chl fluorescence curves were normalized using the StatSoft, Inc. computer program (2014) STATISTICA (data analysis software system, version 12, www.statsoft.com).

Conclusions
Green alders growing in positions below the upper forest line generally show better physiological parameters than plants growing in higher localities which is associated with less environmental stress. Thanks to its multiple physiological attributes, such as efficient vegetative reproduction and the ability to perform symbiotic nitrogen fixation, A. viridis can rapidly take over new habitats, displacing indigenous plant species from lowland ecosystems. The natural occurrence of this taxon in BNP is above the upper limit of the forest. However, recently it has occupied lower positions, mostly in secondary post-farming areas. The growing range of A. viridis to the north and west indicates that this species may become extensive in Poland.