Dynamics of Isomeric and Enantiomeric Fractions of Pinene in Essential Oil of Picea abies Annual Needles during Growing Season

Norway spruce (Picea abies (L.) H. Karst.) is one of the most important commercial tree species distributed naturally in the Boreal and subalpine forest zone of Europe. All parts of spruce trees, including needles, accumulate essential oils with a variety of chemical properties and ecological functions, such as modulating plant–insect communication. Annual needle samples from 15 trees (five from each of three habitats) of 15–17 years old were assayed for essential oils and their major compounds, including α-pinene, β-pinene, (1S)-(−)-α-pinene, and (1R)-(+)-α-pinene across a growing season. Results showed strong positive correlation between percentages of α- and β-pinene isomers (r = 0.69, p < 0.05) and between pinene isomers and essential oils: α-pinene correlated with essential oil stronger (r = 0.62, p < 0.05) than β-pinene (r = 0.33, p < 0.05). Correlation analyses performed with some weather conditions, including average monthly temperature, growing sum of effective temperatures over 5 °C, duration of sunshine, accumulated precipitation, relative humidity, and pressure, showed that temperature is the most important weather condition related to pinene dynamics: negative correlations of moderate strength were established between percentages of α- and β- pinenes and average monthly temperatures (r = −0.36, p < 0.01, n = 75 and r = −0.33, p < 0.01, n = 75, respectively). Out of pinene enantiomers, only (1S)-(−)-α-pinene showed some negative correlation with monthly temperature (r = −0.26, p < 0.05, n = 75). Different patterns of essential oil and pinene dynamics during growing season within separate habitats suggested that some genetic variables of Picea abies might be involved.


Introduction
Norway spruce, Picea abies (L.) H. Karst. (Pinaceae), is a large essential oil-bearing evergreen coniferous tree species valued largely for timber production. It is the main species in the Boreal and subalpine conifer forests of Europe, from Central (in mountains) to Northern and Eastern Europe, up to the Ural Mountains [1].
Picea abies accumulates essential oils in all parts of the plant-in leaves (needles), wood, cones, and bark [2]. However, certain peculiarities exist regarding the quantitative and qualitative composition and distribution of essential oils across various parts of the plant. For example, it was reported that in Picea abies stem wood, higher percentages of α-pinene accumulate in heartwood than in sapwood, while β-pinene takes over in sapwood, compared to heartwood [3,4]. When comparing the yield of oil extractives from wood and bark of Picea abies, it was established that bark oil content exceeds that of wood by nearly three times [5]. A study on cone essential oil composition reported that α-pinene comprised 11%, and β-pinene amounted to 30% of the relative constitution of monoterpene fraction [6]. Monoterpenes constitute the biggest part of Picea abies essential oils, amounting to about 70%, with limonene, camphene, and pinene dominating often [7][8][9]. Most of the monoterpenes are good information carriers over long distances because they present some plants, maximum yearly amounts were recorded during the first month from the start of bud growth with other maximums established in June and July [10]. Similar patterns of seasonal variations in amounts of most terpenes have been reported on white spruce (p. glauca) and blue spruce (p. pungens) [29,30]. A significant reduction in β-pinene, limonene, and α-phellandrene contents along with needle age from one to four years old has been reported for Picea abies [7]. In general, the terpenes in needles of Picea abies are controlled by genetic, physiological, and external factors. The most important factors seem to be genetic control, causing great differences between wild trees even at the same narrow location, and needle age, particularly during development [8].
The aim of this study was to establish seasonal dynamics of the quantitative composition of essential oils with the emphasis on αand β-pinene and enantiomers of α-pinene in annual needles of Picea abies. To achieve this aim, we analyzed monthly changes in percentages of essential oils and pinene compounds in the same age (15-17 years old) trees in three very similar habitats, considered as replications. Correlation analyses between essential oil percentages and various weather conditions were employed to reveal this kind of relationship in seasonal dynamics of needle essential oils in Picea abies. The variation in the essential oil composition, particularly the contents of pinene isomers, is important to predict their biological effects, for example, to determine the most favorable or unfavorable periods for the associated insect pests.

Dynamics of Air-Dry Mass Yield and Quantitative Composition of Essential Oil in Annual Needles during the Growing Season
The dynamics of air-dry mass yield of annual shoots (n = 15 trees) during the growing season is demonstrated in Figure 1a and Table S1. The highest mass yield in August was 2.5 times bigger than the lowest one in May. The highest leap of growth was observed in the period from May to June (the mass has nearly doubled), while the lowest (only 2%) occurred in the period from June to July. A very small decrease was observed at the end of the growing season, from August to September. Most of the mean values of the mass yield differed significantly (p < 0.05) between months, except for those between June and July and between August and September.
Molecules 2021, 26, 2138 4 of 16 in this tree in July was significantly lower (p < 0.05) than in any other four months. Oneway ANOVA revealed no significant differences between means of essential oil percentages between habitats across months except between habitats 1 and 3 (p = 0.048) in May. However, no significant effect of habitat was detected (p = 0.119). A bit different picture was obtained when pooled monthly data were compared between habitats, i.e., the significant differences were established between habitats 1 and 2 (p = 0.045) and 1 and 3 (p = 0.003). An estimated habitat effect in this case amounted to 11.8% (F = 4.825, p = 0.011).
(a) (b) Figure 1. Dynamics of air-dry mass yield (a) and essential oil percentage (b) in Picea abies (n = 15) needles during the growing season. Error bars indicate ± SD; different letters denote statistically significant (p < 0.05) differences. Percentages of essential oils varied with a range of 0.41-0.52% (v/w) on average during the growing season (Figure 1b, Table S1). However, no statistically significant differences were observed between months. The lowest average (0.41 ± 0.18% (v/w)) of essential oil percentage was established at the beginning of the growing season and the highest in July (0.52 ± 0.27% (v/w)). A decrease in essential oil percentage was observed in the second half of the growing season. However, Picea abies demonstrated different variation patterns of the quantitative composition of essential oils during the growing season, measured by the coefficient of variation (from CV = 4% in tree 1.1 to CV = 55% in tree 3.3), and percentages of essential oils differed significantly (p < 0.05) between months in most individuals of Picea abies (Figure 2a-c). For example, the highest percentages of essential oils were established in May, in June, in July, and in August in three, six, four, and two trees, respectively. Individuals No. 2.2 and No. 3.3 distinguished most, showing the highest percentages of essential oils (1.22 ± 0.04% (v/w) in June and 1.04 ± 0.02% (v/w) in July, respectively), significantly (p < 0.05), differing from the percentages in other months ( Figure 2). While the lowest percentages of essential oils were established mostly at the beginning of the growing season (40% of trees in May), July was the month with the lowest percentage of essential oils in Picea abies No. 3.4 (Figure 2c). The amount of essential oils accumulated in this tree in July was significantly lower (p < 0.05) than in any other four months. One-way ANOVA revealed no significant differences between means of essential oil percentages between habitats across months except between habitats 1 and 3 (p = 0.048) in May. However, no significant effect of habitat was detected (p = 0.119). A bit different picture was obtained when pooled monthly data were compared between habitats, i.e., the significant differences were established between habitats 1 and 2 (p = 0.045) and 1 and 3 (p = 0.003). An estimated habitat effect in this case amounted to 11.8% (F = 4.825, p = 0.011). Error bars indicate ± SD; different letters denote statistically significant (p < 0.05) differences between months within plants.

Dynamics of Percentages of Pinene Isomers in Annual Needles during the Growing Season
Variations of α-and β-pinene percentages in essential oils of annual needles of all studied Picea abies trees are presented in Table S2. The results of correlation analysis showed a strong positive correlation between percentages of pinene isomers (r = 0.69, p < 0.05), implying that the syntheses of α-and β-isomers of pinene are quite closely linked . On X-axis, the first part of a combined number indicates habitat number, and the second indicates individual plant number within that habitat. Error bars indicate ± SD; different letters denote statistically significant (p < 0.05) differences between months within plants.

Dynamics of Percentages of Pinene Isomers in Annual Needles during the Growing Season
Variations of αand β-pinene percentages in essential oils of annual needles of all studied Picea abies trees are presented in Table S2. The results of correlation analysis showed a strong positive correlation between percentages of pinene isomers (r = 0.69, p < 0.05), implying that the syntheses of αand βisomers of pinene are quite closely linked to each other in Picea abies. Additionally, statistically significant positive correlations were established between percentages of pinene isomers and essential oils in Picea abies needles with α-pinene correlating to essential oil stronger (r = 0.62, p < 0.05) than β-pinene (r = 0.33, p < 0.05).
The highest amounts of pinenes in annual needles of most investigated trees were found in May; only in three (2.1, 3.2, and 3.3) and two (1.2 and 3.3) individuals, the highest percentages of α-pinene and β-pinene, respectively, were found in other months (Table S2). The dynamics of mean percentages of both pinene isomers of the studied trees (n = 15) demonstrated two different trends (Figure 3a,b). The first trend-May through Juneshowed a very strong decrease in percentages of αand β-pinenes: α-pinene decreased by 1.8 times (p = 0.009) and β-pinene by 2.1 times (p = 0.007). The second trend-June through August-demonstrated a statistically not significant increase in pinene isomers, i.e., by 12% for α-pinene and 18% for β-pinene (Figure 3a
The highest and the lowest percentages of (1S)-(−)-α-pinene in different Picea abies trees were established in different months, but two-thirds of the studied trees demonstrated their highest amounts of (1S)-(−)-α-pinene in May (Table S3). The dynamics of Figure 3. Dynamics of mean α-pinene (a) and β-pinene (b) percentages in essential oils of Picea abies (n = 15) needles during the growing season. Error bars indicate ± SD; different letters denote statistically significant (p < 0.05) differences.

Correlations with Weather Conditions
Pearson correlation analysis was employed to reveal correlations between the dynamics of the studied properties and changes of weather conditions of Picea abies growth environment. The results of correlation analysis with temperature are summarized in Table 2 and data on weather conditions are presented in Table 3.   Error bars indicate ± SD; different letters denote statistically significant (p < 0.05) differences.

Correlations with Weather Conditions
Pearson correlation analysis was employed to reveal correlations between the dynamics of the studied properties and changes of weather conditions of Picea abies growth environment. The results of correlation analysis with temperature are summarized in Table 2 and data on weather conditions are presented in Table 3.   Strong correlation was established between air-dry mass yield and sum of effective temperatures (≥5 • C) (r = 0.86, p < 0.01, n = 75). No correlations were established between the latter indicator and any those of essential oils, except for weak negative correlations with percentages of α-pinene (r = −0.286, p < 0.05, n = 75) and β-pinene (r = −0.252, p < 0.05, n = 75). Additionally, no correlation was established between air-dry mass yield and essential oil percentage during the growing season. Some weak correlations were established between the monthly accumulated precipitation and percentages of α-pinene (r = 0.313, p < 0.01, n = 75) and β-pinene (r = 0.295, p = 0.01, n = 75), of which the former was best pronounced within the habitat No. 1, where α-pinene correlated with precipitation quite strongly (r = 0.525, p < 0.01, n = 25). No correlations were established between the studied Picea abies properties and other weather conditions, such as humidity, pressure, and sunshine duration.

Discussion
The growing season of Picea abies in Lithuania starts in May and ends in September-October, depending on autumn weather conditions [35,36]. Our observations showed a continuous growth of annual shoots during the growing period. However, nonmonotonic mass change revealed the pattern of growth rate from May to August, with the fastest growth rate in May through June and a slightly decreasing growth rate from August to September (Figure 1; Table S1). Correlation analysis revealed a strong linear relationship between air-dry mass yield and average monthly temperature during the growing season (r = 0.77, p < 0.01, n = 75) ( Table 2). Additionally, even stronger correlation was established between air-dry mass yield and sum of effective temperatures (≥5 • C) (r = 0.86, p < 0.01, n = 75). Only weak negative correlations established between sum of effective temperatures and percentages of α-pinene (r = −0.286, p < 0.05, n = 75) and β-pinene (r = −0.252, p < 0.05, n = 75), suggesting that this derivative meteorological indicator concedes to the use of temperature, a directly measurable factor. The absence of correlation between air-dry mass yield and essential oil percentage during the growing season may have some implications regarding the use of Picea abies needles as raw material for essential oil production. Regarding habitat effect on essential oil percentages, significant differences were established between habitats 1 and 2 (p = 0.045) and 1 and 3 (p = 0.003), suggesting that more fertile soil in habitat 1 could have discerned it out of the other two sites. However, an estimated habitat effect, in this case, amounted only to 11.8% (p = 0.011), which does not permit us to make any deeper inferences. Essential oil percentages also demonstrated variation during the growing season-the lowest average of essential oil percentage was established in May (0.41%) and the highest in July (0.52%). These data correspond to those obtained in the geographical neighborhood of Lithuania-the percentage of essential oils in needles of Picea abies growing wild in Latvia varied from 0.36% to 0.53% and was lower in April-May [37,38]. However, the variation of essential oil percentages in individual trees during the growing season showed different patterns. The lowest percentages prevailed in the individual 3.5 (all months except May) and the highest percentages in the individual 2.2 (except for May and July) (Table S1). Thus, no significant correlation between essential oil percentage and the average monthly temperature was established ( Table 2). Although different environmental conditions (climate, light conditions, soil chemistry) and harvesting time of raw material can influence the quantitative composition of essential oils [39], biosynthesis of volatile oils is a genetically controlled process. In our study, all individuals within a habitat grew under the same environmental conditions and thus were similarly influenced regarding their biosynthesis of essential oils. Therefore, some genetic variables of the studied trees might better explain differences in dynamics of essential oil percentages between them during the growing season. Some researchers argue that terpene composition varies less than the absolute amounts, thus allowing to discern Picea abies genotypes based on the variability of relative amounts of certain terpenes [10]. Based on relative amounts of limonene and myrcene, they distinguished two genotypic groups, namely, limonene > myrcene (group 1) and myrcene ≥ limonene (group 2). In our study, all 15 spruce trees had limonene dominating over myrcene, from 8 (tree 2.2) to 35.3 (tree 2.5) times on average (Table S4). Interestingly, both extreme limonene/myrcene ratios and two next to the extremes (trees 2.4 and 2.3) occurred in trees from the same habitat No. 2. Nevertheless, according to Schönwitz et al. (1989), only one genotype could be distinguished based on the limonene to myrcene ratio [19]. According to the researchers who studied odors of Norway spruce seedlings [40], spruce was defined as a bornyl acetate chemotype when its emission consisted of 35-54% bornyl acetate; 13-29% of limonene, and the content of every other compound was less than 10%. Only one tree (No. 2.1) met these conditions in our study, with 39.42% of bornyl acetate and 15.56% of limonene contents (Table S4).
The highest percentages of pinene stereoisomers αand βreached 9.32% and 3.11%, respectively (Table S2). Literature data suggest that percentage of α-pinene can vary from 14.2 to 27.7% in essential oils of needles of this conifer [2,31]. Other authors indicate that α-pinene can average to 10.67% of essential oil of Picea abies needles [41]. Meanwhile, β-pinene always has been found in amounts several times lower than those of α-pinene [31,41], although both isomers usually occur together in essential oils [2]. Our study demonstrated that an increase in the percentage of α-pinene was accompanied by an increase of β-pinene in essential oils of annual needles of Picea abies (r = 0.69, p < 0.05). Similar correlation patterns between these pinene isomers were also observed in cones and leaves of Juniperus communis with r = 0.44 (p < 0.05) and r = 0.50 (p < 0.05), respectively [42]. Therefore, it can be presumed that the syntheses of αand βisomers of this monoterpene are mutually dependent in Picea abies and in other conifers.
The highest amounts of pinenes in annual needles in May were followed by a sharp decrease in June (Figure 3). Negative correlations of moderate strength were established between the percentages of αand βpinenes and average monthly temperatures (r = −0.36, p < 0.01, n = 75 and r = −0.33, p < 0.01, n = 75, respectively) ( Table 2). The highest percentages of α-pinene and β-pinene in needles of Picea abies growing in Latvia during this period (May-September) also were established in May (0.67% and 0.47%, respectively); however, the amounts of both isomers went down gradually, reached the lowest percentages (0.39% and 0.15%, respectively) only in July, and began to rise again just in August [38]. Similar consistent patterns have been found by the studies on p. sitchensis and p. pungenspercentages of pinenes rose in essential oils of needles in May and then declined during the growing season [30,43].
According to the Lithuanian dendrologists, the growth of new Picea abies shoots in Lithuania starts in May [36,44]. A widely accepted concept exists that the onset of the growing season of Picea abies estimated by bud burst depends on the growing degree days above a certain threshold temperature, usually 5 • C [45][46][47][48]. However, one of the latest studies showed that buds of three spruce species (including Picea abies) were sensitive to frost probability for early phenological stages, whereas growing degree days controlled the remaining stages [49]. Picea abies, especially young 15-to 20-year-old trees, suffer from larvae of needle-eating pest Lygaeonematus abietinus Christ. (Tenthredinidae) almost every year, usually in May [36,50]. The larvae of L. monacha are also active in May when the air temperature reaches 10-15 • C [51,52]. Thus, the sprouting time of Picea abies needles coincides with the larvae stage of needle-eating pests L. abietinus and L. monacha. Most of the terpenes found in essential oils of plants have been demonstrated to possess insecticidal and repellent activities [53]. Although less research data is available on the effects of pinene as a repellent, compared to many other terpenes, this bicyclic monoterpene has also been shown to repel Aedes aegypti, Aedes albopictus, and Musca domestica [54][55][56]. Therefore, noticeably higher percentages of pinene in essential oils of annual needles at the beginning of the growing season, compared to the further growth period, may relate to plant defense against L. abietinus larvae. A study on the effects of Ceratocystis polonica on Picea abies showed significantly higher contents of terpenes in less susceptible and less damaged trees, compared to those more damaged 30 days after the inoculation [57]. Therefore, the increased amount of diterpenes can be considered a defensive response to the fungal infection. This may also indicate a possible relationship between the monoterpene contents and resistance of Picea abies to some biotic stress. As reported in one of the reviewed sources, needles of healthier-looking trees contain higher amounts of monoterpenes [7].
In our study, the highest percentage of (1S)-(-)-α-pinene was established in May (Table S3). It was reported that the enantiomeric composition in Picea abies change from more (−)α-pinene in the younger parts to more (+) enantiomer in the older parts [22]. In our study, higher percentages of (1S)-(-)-α-pinene in May, in comparison to other months, correspond to the reported observation. Regardless of the negative correlation between both enantiomers of α-pinene, the lowest percentage of (1S)-(-)-α-pinene and the highest percentage of (1R)-(+)-α-pinene were observed in different months. It could be related to the established fact that monoterpene-producing enzymes (i.e., monoterpene synthases) are enantiomer specific [62]. Effects of environmental factors could not be excluded as well. For example, it has been reported that biosynthesis of (−)-α-pinene is dependent on light levels, which either activate the synthase enzymes and/or provide the necessary energy for biosynthesis [60]. However, in our study, correlation analyses between monthly sunshine duration and amounts of any of pinene enantiomers, isomers, or total essential oils revealed no correlation. This result could be influenced by daily photoperiodism and the angle of sunrise. For example, the longest sunshine duration, 350 h, occurred in August (Table 3), when days were shorter, compared to June by more than 2 h on average, and sunrise was correspondingly lower than its zenith in June. Some weak correlations were established only between the monthly accumulated precipitation and percentages of α-pinene (r = 0.313, p < 0.01, n = 75) and β-pinene (r = 0.295, p = 0.01, n = 75), of which the former was best pronounced within the habitat No.1, where α-pinene correlated with precipitation quite strongly (r = 0.525, p < 0.01, n = 25).
Our results show that different essential oil chemotypes of Picea abies could be found under different growing and environmental conditions. Different chemical compositions can considerably affect the biological activity of the essential oil, including the antimicrobial activity, as revealed with thyme (Thymus vulgaris L.) extracts [63,64]. It is evident that a better knowledge of seasonal dynamics of pinene isomers and, particularly their chiral compounds, is increasingly important in revealing insect-plant or plant-plant communication in future warmer climate conditions since insects and plants discriminate and respond to changes in chiral compounds [60]. Our study showed that both intrinsic and environmental factors predefine the dynamics of isomeric and enantiomeric fractions of pinene in essential oils of Picea abies annual needles during the growing season. Different patterns of essential oil dynamics during the growing season within separate habitats suggested that some genetic variables of Picea abies might be involved. Therefore, further research is needed to justify factors causing chemotypic variation in Picea abies better. This could be best achieved by extending the studies on the chemical composition of natural extracts and their biological activities, including natural antioxidants and antimicrobial compounds [65].

Plant Material and Habitats
Three Picea abies habitats of anthropogenic origin were chosen in the Vilnius region (Table 4).  [35]. Annual shoots were collected at a height of 1.5-2.5 m from all sides of each plant. Shoot samples were then dried at room temperature separately. To establish air-dry mass, shoots were weighed before and after drying. The yield of air-dry mass was then calculated by the following formula: Y = C/D*100%, where Y-air-dry mass yield (%); C-mass of air-dry shoots (g), D-mass of fresh shoots (g).

Isolation and Analysis of Essential Oils
Essential oils were isolated from air-dry needles (free from stems) by hydrodistillation in Clevenger-type apparatus (Simax) for two hours [66]. Distillation of essential oils of each sample was carried out in three replications, and the essential oil of the same sample was collected in the same bottle. The content of essential oil was calculated in % (v/w), based on the weight of air-dry mass of needles.
Afterward, 1% solutions of essential oils were prepared with a mixture of diethyl ether and n-pentane (1:1) for further investigations. The analysis of essential oils was carried out using a FOCUS GC (Thermo Scientific, Waltham, MA, USA) gas chromatograph with a flame ionization detector (FID). Data were processed with the CHROM-CARD S/W. The silica capillary column TR-5 (30 m, i. d. 0.25 mm, film thickness 0.25 µm) was used for the analysis of α and β pinenes, myrcene, limonene, camphene, and bornyl acetate with the following GC parameters: carrier gas helium flow rate 1.6 mL/min; temperature program from 40 • C to 250 • C increasing at 4 • C/min; detector temperature 260 • C; split injector was heated at 250 • C. The identification of α and β pinenes, myrcene, limonene, camphene, and bornyl acetate was carried out by the comparison of the retention time (RT) of their GC peaks in the FID chromatograms with the RT of α and β pinenes, myrcene, limonene, camphene, and bornyl acetate analytical standards (Sigma-Aldrich, St. Louis, MI, USA) under the same GC parameters and column. The percentage amounts of these compounds were recalculated according to the areas of the FID chromatographic peaks assuming that all constituents of the essential oil comprise 100%.

Meteorological Data
Weather data for the Vilnius region were obtained from the Lithuanian Hydrometeorological Service (sunshine duration, accumulated precipitation, and mean daily temperatures for the sum of effective temperatures) and from Timeanddate.com [67] ( Table 3).

Statistical Analysis
Statistical data processing, including calculation of means, standard errors, standard deviations, coefficients of variation (CV), Spearman's rank correlation coefficients (r), the Kruskal-Wallis test, and probabilities (p), was carried out with STATISTICA ® 7 and MS Excel. One-way ANOVA with Kruskal-Wallis test was used for the assessment of quantitative changes of essential oils in individual plants during the growing season. Pearson correlation analyses were performed with SPSS Statistics and ANOVA (comparison of means and GLM univariate analysis) used for statistical evaluation of habitat effects.

Supplementary Materials:
The following are available online, Table S1: Air-dry mass yield of annual shoots and contents of essential oils in leaves of Picea abies observed in May through September. In Plant # column, the first numeral stands for habitat number (as described in Materials and Methods), and the second stands for individual plant number within that habitat (as in Figure 2); SD -standard deviation; CV -coefficient of variation, Table S2: Variation of percentages of pinene isomers in the essential oil of Picea abies leaves observed in May through September. In Plant # column, the first numeral stands for habitat number (as described in Materials and Methods), and the second stands for individual plant number within that habitat (as in Figure 2); SD -standard deviation; CV -coefficient of variation, Table S3: Variation of ratio percentages of (1S)-(-) and (1R)-(+) enantiomers in the α-pinene fraction of Picea abies leaves observed in May through September. In Plant # column, the first numeral stands for habitat number (as described in Materials and Methods), and the second stands for individual plant number within that habitat (as in Figure 2); SD -standard deviation; CV -coefficient of variation, Table S4: Variation of myrcene, limonene, camphene, and bornyl acetate percentages in the essential oil of Picea abies leaves during the growing season, May-September. In Plant # column, the first numeral stands for habitat number (as described in Materials and Methods), and the second stands for individual plant number within that habitat (as in Figure 2); SD -standard deviation, Figure S1: Chiral-phase capillary GC analysis of (1S)- ( Data Availability Statement: Supplemental data are provided in Tables S1-S4 and Figure S1.