Emission of Fire-Promoting Volatiles from Picea omorika (Pančić) Purk Needles in Different Forest Communities

Abstract

The importance of studying the role of volatiles in flammability is well recognized, but the relationship between specific compounds and components of flammability is underestimated. In this study, volatiles emitted from Picea omorika (Pančić) Purk. needles were identified and quantified, and the relationship between volatile emission and moisture content, flammability characteristics and bioclimatic coefficients was investigated. Fresh needles from four different forest communities were analyzed for specific volatile organic compounds (using a gas chromatography-surface acoustic wave analyzer, GC-SAW), moisture content (based on fresh and oven-dried mass), and flammability components (using an epiradiator). Five monoterpenes, one sesquiterpene, and one alcohol were identified and their amounts, moisture content, and flammability components differed between the populations. The amounts of myrcene and β-pinene correlated significantly with moisture content and time to ignition. Myrcene content also correlated with flame duration, site altitude, mean annual temperature, variation in mean annual temperature, and the Ellenberg quotient. The emission of myrcene correlates with ecophysiological, flammability, topographic, and climatic variables. This suggests that myrcene-rich plant species may be of particular interest for research into flammability, especially from the point of view of rapid ignition.

1. Introduction

Plant flammability is a highly complex functional trait associated with a variety of morphological, ecophysiological, and chemical traits [1]. While various plant traits have been extensively studied, the role of specialized metabolites (historically referred to as secondary metabolites) in flammability is still underestimated. In contrast to primary metabolites, which are common to all living organisms and required for vital biological functions, specialized metabolites are usually species-, organ-, and tissue-specific [2]. Conifers synthesize a variety of specialized metabolites (mainly terpenes) whose composition can vary between and within species [3]. These compounds can be stored in specific organs such as resin ducts or resin blisters (where they apparently fulfill a defense function against pathogens or herbivores) or emitted immediately after their biosynthesis [4]. As conifers produce, store, and release large amounts of specialized metabolites, their volatiles dominate in forest communities [5]. The emission of these volatiles is generally associated with plant communication and their response to biotic stimuli [6]. The most common temperate conifer species (genera Pinus, Picea, and Abies) usually store terpenes in constitutive resin, but can also emit isoprene or de novo produced monoterpenes [7]. The constitutive mixture of terpenes produced by secretory epithelial cells and stored in resin cells [8] can be released as biogenic volatile organic compounds (BVOCs). In general, the terpene concentration in plants is about 1%–2% of their dry weight, but in some species it can increase to 15%–20% of dry weight [4]. Coniferous forests are the most important emission sources of biogenic volatile compounds worldwide [9] and produce considerable amounts of terpenes that interact with other compounds in the atmosphere and enter biogeochemical cycles, affecting air quality and local climate.

Since the first indications that plant volatiles affect flammability [10,11], numerous empirical studies have provided data justifying their inclusion in flammability studies, e.g., [12,13,14,15,16,17]. This need is becoming increasingly important in light of future climate change, which predicts a drier and warmer environment that will lead to increased production and emission of VOCs. Further research should also shed light on the still incompletely understood role of the different VOC classes. Previous studies have found divergent correlations between certain compounds and flammability phases [18,19] or reported their lesser importance compared to moisture content [20,21]. Studies suggest that there is a relationship between moisture content, VOC emissions and flammability [22,23]. However, the complex relationship between moisture content, volatiles, flammability and climatic factors is still unexplored.

In plant flammability studies, volatiles are usually sampled using extraction or capture techniques, which can lead to losses of thermolabile components, chemical transformations and discrepancies in the accurate representation of VOC content [24]. Although newer methods such as headspace analysis with an electronic analyzer compensate for the disadvantages of multi-step sample preparation, they have not yet become the standard for VOC characterization in flammability research. The development of electronic noses (e-noses) represents a significant advance in the rapid and straightforward analysis of VOCs. Despite certain limitations, such as the short column length, the stability of e-noses has proven to be satisfactory under optimal operating conditions and with a predefined calibration of standard components [25,26].

Picea omorika (Pančić) Purk., commonly known as Serbian spruce, is a Tertiary relict species on the Red List [27] whose natural range is restricted to western Serbia and eastern Bosnia and Herzegovina. Previous studies have shown that forest fires have significantly affected its historical and current distribution [28,29]. The increased occurrence of fires associated with climate change fuels the great interest in studying the fire-related characteristics of this species [30]. The aim of this study was to determine VOC emissions in natural P. omorika populations and their relationship to flammability traits.

2. Materials and Methods

2.1. Study Site

The study was conducted in Tara National Park, a region of global ecological importance that covers most of the Tara Mountains in western Serbia (43°57′ N, 19°28′ E). The Tara Mountains, with an average altitude of 1000 to 1200 m, have a climate with mild to cool summers and severe cold winters, with an average annual temperature of 7.9 °C. Temperature fluctuations throughout the year are minimal, with an average temperature of 4.5 °C in January and 16.7 °C in August. The average annual precipitation is 977 mm and the humidity remains consistently high at 83.4%. The forest ecosystems in Tara National Park are mosaic-like and consist of mixed and coniferous forests. The Serbian spruce populations live at altitudes between 700 and 1600 m, typically in steep limestone cliffs or fire pits, often coexisting with spruce, fir, white and black pine, beech, and other deciduous trees.

2.2. Plant Material

The study was conducted in August 2022, a month with the highest temperatures, the lowest precipitation, and an increased incidence of forest fires based on long-term observations. The plant material was collected according to the protocol described by Cornelissen et al. [31]. The needles of P. omorika were collected from three natural populations: (I) Zmajevački potok, (II) Bilo, and (III) Kanjon Brusnice, and an even-aged plantation of Picea abies L. Karst. and P. omorika: (IV) Osluša.

Table 1. Vegetation diversity, position, and basic climate characteristics of studied sites.

Location	Composition of Forest Community	P. omorika Contribution (%)	Altitude (m)	MAT 1 (°C)	chMAT 2 (°C)	EQ 3 (°C/mm)	Coordintes
I Zmajevački potok	Picetum—Omorikae Excelsae pinetosum mixtum, dominated by P. abies and P. omorika, also present Abies alba, Pinus nigra, P. sylvestris, Betula pendula, Sorbus aucuparia.	34	800	6.9	1.93	18.0	43°51′30″ 19°25′35″
II Bilo	Picetum—Omorikae abietis calcicolum, where P. omorika dominates and forms monospecies stands, and P. abies is present in much lesser extent.	80	1200	8.2	1.81	20.7	43°55′19″ 19°20′08″
III Kanjon Brusnice	Erico—Picetum omorikae mixtum, mixed community with P. abies, P. omorika, P. nigra, Carpinus betulus, Betula pendula, Populus sp., Abies alba, Fraxinus nigra, Acer pseudoplatanus, Erica carnea.	9	900	7.8	1.80	19.5	43°55′44″ 19°17′05″
IV Osluša	Plantation of P. abies and P. omorika.	25	1000	8.7	1.79	21.7	43°56′50″ 19°28′55″

¹ Mean annual temperature for the period 1961–2020; ² Change in mean annual temperature for the period 2011–2020 compared to the period 1961–1990; ³ Ellenberg’s climate quotient represents the mean temperature of the warmest month (°C) divided by the annual sum precipitation (mm) and multiplied by 1000.

shows the geographical location, site characteristics, and vegetation characteristics of the studied populations. The climate data for the studied sites for the period 1961–2020 were downloaded from the Digital Climate Atlas of Serbia [32].

Seven well-grown tree specimens were selected from each stand, free of pathogens and herbivore damage and spaced 50 m apart. Five outer branches were harvested from each tree at a height of about 4 m. After harvesting, three twigs with two-year-old needles from the same branch were placed in sealed plastic bags and stored in an ice box until the measurements were taken the next day (one each for measuring the moisture content of the leaves, the chemical analyses and the flammability test).

2.3. Leaf Moisture Content

The needles were measured for fresh weight (per population: 7 trees × 5 branches × 5 needles) and dry weight (oven-dried at 75 °C until a constant weight was reached). The moisture content (MC) was calculated following the equation:

$$\mathrm{MC}={\displaystyle \frac{\mathrm{Mf}-\mathrm{Md}}{\mathrm{M}\mathrm{f}}}\times 100,$$

where Mf represents the fresh mass and Md represents the dry mass.

2.4. GC-SAW Volatile Analysis

Volatiles were analyzed using an electronic nose GC-SAW (zNose™ 4300, Electronic Sensor Technology, Newbury Park, CA, USA) equipped with a DB-5 column (1 m length, 0.25 mm i.d., 0.25 µm film thickness). Approximately 1 g samples of needles were sealed at room temperature in 40 mL screw-cap glass vials with silicone/PTFE septa and stored at room temperature for three hours prior to headspace sampling. During sampling, headspace vapor was aspirated into the e-nose through a bent needle connected to the inlet nozzle. Details of the GC-SAW equipment and the analysis are described in Du et al. [33]. The following chromatographic conditions were applied: injection time of 20 s, inlet temperature of 200 °C, valve temperature of 165 °C, detector temperature of 60 °C, column ramp from 40 to 200 °C at 3.0 °C/s, helium flow rate of 3.0 mL/s, and data acquisition time of 55 s.

zNose™ is a highly sensitive instrument that can detect volatile substances in parts-per-billion. Initially, good chromatographic conditions for the volatiles of P. omorika were found by trial and error to separate the compounds. The most conspicuous chromatographic peaks were selected for further analysis. Their identification was performed by matching Kovats’ standardized retention index (RI) values (using standard n-alkanes of C₈-C₂₀ in hexane, supplied by Sigma-Aldrich, St. Louis, MA, USA) with the literature data for the same column [34] and for P. omorika [35]. Peak identities were confirmed with standards run under the same chromatographic conditions. Standards dissolved in hexane (cis-3-hexen-1-ol, α-pinene, β-pinene, myrcene, limonene, bornyl acetate, trans-caryophyllene, all purchased from Sigma-Aldrich, St. Louis, MA, USA) were injected directly into the e-nose. The concentration of volatiles was determined by establishing calibration curves using the external standard method. The calibration curves of the standard compounds with R² > 0.99 were linear in the concentration range found in the plant samples. The results are expressed in µg/g fresh weight.

2.5. Flammability Testing

Time to ignition (TTI) and flame duration (FD) for fresh foliage were determined using a 500 W epiradiator (model 534Rc2, Quartz Alliance SAS, Villemer, France) with a 10 cm radiant disk and a nominal surface temperature of 420 °C in a controlled environment. The horizontally oriented epiradiator had a pilot flame positioned 4 cm above the center of the disk [36]. Once the device reached a steady state, the samples were evenly distributed on the radiant disk. A total of 28 samples, each containing 3.0 ± 0.1 g of plant material, were analyzed consecutively [37]. The duration between the placement of a sample and the ignition of the flame (TTI) and the duration until the flame was extinguished (FD) were recorded using a digital timer.

2.6. Statistical Analysis

Four populations of P. omorika were compared on the basis of ten measured variables: MC, VOCs (cis-3-hexen-1-ol, α-pinene, β-pinene, myrcene, limonene, bornyl acetate, trans-caryophyllene), TTI and FD. The Shapiro–Wilk test was used to test normality assumptions; p > 0.05 indicated normal distribution. ANOVA or the Kruskal–Wallis rank sum test were used to assess differences between populations. To control for type I error rate in hypothesis testing, the false discovery rate (FDR) p-adjustment method was used. Pearson correlations were used to assess the relationships between the measured variables and between the measured variables and the bioclimatic factors. Principal component analysis (PCA) was used to visualize the data structure. Statistical analyses were performed using the statistical software R, version 4.2.1 [38].

3. Results

The most abundant volatiles in all samples were cis-3-hexen-1-ol and β-pinene (Figure 1). The populations differed in TTI, FD, MC, and in the emission of cis-3-hexen-1-ol, myrcene, and trans-caryophyllene. The differences in the concentrations of α-pinene, β-pinene, limonene, and bornyl acetate emitted from the needles were not significant between the populations.

The relationships between the measured variables and between the variables and the bioclimatic coefficients were quantified using Pearson’s correlation (

Table 2. Pearson’s correlation coefficients between the measured variables and between the variables and the bioclimatic coefficients. Statistically significant correlations are shown in bold. Asterisks denote statistical significance: * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001.

	cis-3-Hexen-1-ol	α-Pinene	β-Pinene	Myrcene	Limonene	Bornyl Acetate	trans-Caryophyllene	TTI	MC	FD
TTI	−0.33	0.16	0.39 *	0.55 **	−0.12	−0.10	0	1
MC	−0.20	0.11	0.40 *	0.48 **	−0.13	−0.18	0.10	0.85 ***	1
FD	0.06	−0.14	−0.32	−0.41 *	0.20	−0.18	−0.09	−0.81 ***	−0.82 ***	1
Alt	0.27	−0.18	−0.45 *	−0.55 **	0.18	−0.03	−0.13	−0.91 ***	−0.93 ***	0.90 ***
MAT	0.37 *	−0.05	−0.18	−0.46 *	0.03	0.45 *	0.30	−0.65 ***	−0.73 ***	0.35
chMAT	−0.16	0.05	0.26	0.41 *	−0.06	−0.31	−0.04	0.66 ***	0.90 ***	−0.57 **
EQ	0.42 *	−0.06	−0.17	−0.48 **	0.03	0.45 *	0.34	−0.66 ***	−0.70 ***	0.34

Abbreviations: TTI—time to ignition; MC—moisture content; FD—flame duration; Alt—altitude; MAT—mean annual temperature for the period 1961–2020; chMAT—change in mean annual temperature for the period 2011–2020 compared to the period 1961–1990; EQ—Ellenberg’s quotient.

). Time to ignition (TTI) showed a positive correlation with both β-pinene (r = 0.39, p ≤ 0.05) and myrcene (r = 0.55, p ≤ 0.01). β-Pinene and myrcene were also significantly correlated with MC (r = 0.4, p ≤ 0.05; r = 0.48, p ≤ 0.01, respectively). Myrcene content showed a negative correlation with the flame duration (FD) (r = −0.41, p ≤ 0.05). Myrcene content, TTI, and MC were significantly correlated with all bioclimatic coefficients.

The PCA resulted in a 3-component model that explains 70% of the variation. The clustering of individuals is visible in the 2D PCA plot (Figure 2). TTI, MC, FD, β-pinene, and myrcene primarily determined the formation of the first axis and influenced the separation of individuals belonging to populations I and II. The variability of limonene, α-pinene, β-pinene, and trans-caryophyllene mainly contributed to the formation of the second axis, while bornyl acetate, cis-3-hexen-1-ol, FD, and trans-caryophyllene were the most important factors for the third axis and the separation of populations III and IV along this axis.

4. Discussion

Although all samples emitted the whole set of studied volatile compounds, the quantities of the individual volatiles varied considerably between populations. The most abundant compounds in all populations were cis-3-hexen-1-ol and β-pinene. The amounts of α-pinene, β-pinene, limonene, and bornyl acetate released from the needles did not differ significantly, while TTI, FD, MC, and emission of cis-3-hexen-1-ol, myrcene, and trans-caryophyllene differed significantly between the populations studied. Population I had the highest content of myrcene, the highest MC, the longest TTI, and the shortest FD. Overall, the levels of volatiles determined were consistent and comparable to those determined by conventional GC-MS methods in P. omorika twigs [35].

The content of β-pinene and myrcene was negatively correlated with altitude, which agrees with the earlier report by Večeřová et al. [39], who found that the total terpene content of P. abies decreases with increasing altitude. The needles that emitted more β-pinene and myrcene had higher TTI values and were less flammable, which is consistent with the findings of Guerrero et al. [16], but can also be attributed to MC, as both β-pinene and myrcene were positively correlated with MC. Both β-pinene and myrcene have antibacterial and antifungal properties [40,41,42,43,44] and play an important role in the activation of signaling pathways during plant defense responses to biotic and abiotic stress factors [45]. Higher MC makes leaves more susceptible to pathogens [46], implying that higher emission of β-pinene and myrcene is one of the defense mechanisms. In addition, the emission of monoterpenes depends on the carbon obtained through photosynthesis, which is limited by water availability [23], implying that lower MC leads to lower emission of monoterpenes. However, it should be noted that the drying treatment of the material can have an influence on the biochemistry and degradability of the cell walls. The MC value can be determined more accurately after withering at 105 °C, which may help to establish more reliable relationships in future experiments. Guerrero et al. [16] suggest that monoterpenes volatilize as highly volatile molecules before ignition occurs and that the measured emissions could be lower at the temperatures used in the epiradiotor experiment (420 °C) due to thermal degradation and/or polymerization. This finding may contribute to a better understanding of our finding that monoterpenes appear to contribute to a decrease in flammability and highlights the difficulty in estimating their influence on the flammability variables.

The thermotolerance hypothesis [47] states that isoprene and certain monoterpenes contribute to the tolerance of leaves to short periods of high temperatures, and their protective effect is attributed to double bonds [48]. Molecules with double bonds are thought to stabilize cell and thylakoid membranes, which become disrupted and leaky during high-temperature bursts, by inserting themselves into and reinforcing the hydrophobic interactions of the membrane bilayer [48]. Of the volatiles investigated in this study, cis-3-hexen-1-ol with one double bond and the acyclic monoterpene myrcene with the 1,3-diene moiety exhibit structures that can contribute to membrane stabilization. The temperature-dependent increase in cis-3-hexen-1-ol emission supports the hypothesis of thermotolerance. On the other hand, myrcene emissions were negatively correlated with the mean annual temperature measured from 1961 to 2020 (MAT), but positively correlated with the increase in mean annual temperature from 2011 to 2020 compared to the period 1961–1990 (chMAT), implying that plants from sites with a larger temperature increase emitted more myrcene.

According to the Ellenberg quotient (EQ) [49], all P. omorika sites examined were humid (low EQ). MAT and EQ were positively correlated with the content of cis-3-hexen-1-ol and bornyl acetate and negatively correlated with the content of myrcene, TTI, and MC. It implies that needles from more humid regions tend to have higher MC and myrcene content, lower cis-3-hexen-1-ol and bornyl acetate content and ignite more slowly. Higher bornyl acetate and cis-3-hexen-1-ol emissions in P. omorika populations from sites with higher MAT and EQ could be a compensatory mechanism for protecting the plant defense system. Temperature fluctuations affect herbivore abundance and activity, and plants have evolved strategies to combat herbivores whose consumption rate and performance increase with temperature [50]. Bornyl acetate has been reported to exhibit antimicrobial activity against a variety of bacterial strains [51,52,53], and cis-3-hexen-1-ol showed antifungal properties [54] and proved to be an important mediator in plant-to-plant communication, facilitating the mobilization of plant stress-related mechanisms against pathogens and herbivores [45,55,56].

Our results are consistent with other studies showing a correlation of essential oil profiles with bioclimatic variables. For example, chemical variability between provenances of Pinus halepensis Mill. was shown to be dependent on bioclimatic temperature indices measured over a 20-year period [57]. Lakušić et al. [58] have also shown that temperature-related bioclimatic indices are the most important abiotic factors influencing the variability of Rosmarinus officinalis L. essential oil composition. On the other hand, in the study conducted by Rajčević et al. [59], the essential oil profile of Pinus peuce Griseb. showed a high correlation with bioclimatic parameters related to altitude, temperature, and precipitation, measured over a 50-year period, while P. heldreichii Christ. showed no correlation with bioclimatic parameters. These studies shed light on the plasticity of individual species and their ability to adapt to the changing environment. Nevertheless, real-time studies on the influence of climate on the chemical and phenotypic variability of plants over time are needed.

The identity of the neighboring plants, i.e., the community structure, has a considerable influence on the chemical defense of the plants [60]. The proportion of P. omorika in the forest communities varied considerably (34%, 80%, 9%, 25% on sites I, II, III, and IV, respectively), and the diversity of trees ranged from only two species (plantation) to more than ten coniferous and deciduous species. It has been shown that the air in forests where conifers predominate has a higher VOC concentration than in forests with more deciduous trees [61]. Plants inhabiting more diverse communities are exposed to simultaneous competitive, herbivore, and pollination pressures that influence the distribution of defense compounds. Consequently, it has been suggested that plant community diversity may also affect VOC emissions, but there is insufficient empirical evidence to support this claim [62].

The warming of the Western Balkans has accelerated, especially since the 1980s. Most of the Serbian territory, including the studied area of the Tara Mountains, falls under the warm temperate climate type—fully humid with warm summers (Cfb type, according to the Köppen−Geiger climate classification). The climate conditions from 1961 to the present show a significant increase in temperature and a change in precipitation patterns, and analyses of future changes according to the Fifth Assessment Report (AR5) of the Intergovernmental Panel on Climate Change (IPCC) indicate that the temperature could increase by 1.9 °C to 4.4 °C (depending on the different emission-related scenarios) by the end of the century, while summer precipitation could decrease [63]. The intra-annual redistribution of precipitation towards longer periods of drought has already led to an increase in the frequency and intensity of heatwaves and subsequently to an increase in forest fires [64].

More frequent and severe droughts already have a significant impact on the functioning of forest ecosystems [65]. Droughts cause phenotypic variation in numerous functional traits of trees, including terpene concentrations, which have been shown to correlate with the severity of drought stress [66]. High ambient temperatures stimulate the activity of enzymes involved in terpene biosynthesis to produce volatile compounds and volatilization itself [7]. The reciprocal relationship between plant volatiles and climate is evident in large forest areas, where they can significantly alter local climatic conditions (formation of light-scattering blue haze) [67].

The growing evidence of the involvement of VOCs in plant flammability suggests that future forestry practice should take their emissions into account and points to a promising path for the use of advanced technologies. Some progress has already been made in this field with the development of an electronic multisensor capable of predicting wildfire onset by detecting the increase in eucalyptol and α-pinene concentrations in forested areas [68]. The compounds identified and quantified in this study have different flash points according to the PubChem database: β-pinene (31 °C), α-pinene (33 °C), myrcene (39.4 °C), limonene (45 °C), cis-3-hexen-1-ol (54 °C), bornyl acetate (88 °C), and trans-caryophyllene (101 °C) [69]. This could indicate that the amount of highly flammable volatiles is important for the rapid ignition of the plant material. Our results add to the current knowledge on the role of volatiles in plant flammability and point the way for future research on the integration of climate, volatiles, and flammability.

5. Conclusions

Different ratios of certain volatile compounds in fresh needles of P. omorika from four forest communities may indicate that biodiversity affects emissions, although more evidence is needed to support this finding. Of all the volatiles identified and quantified in this study, the emission of myrcene correlates with ecophysiological, flammability, topographic and climatic variables (moisture content, time to ignition, flame duration, site altitude, mean annual temperature, change in mean annual temperature from 2011 to 2020 and Ellenberg’s climate quotient). This suggests that further research into the relationship between vegetation and fire should pay particular attention to myrcene-rich plant species and their suitability for reforestation programs. Analyzing the relationships between volatiles and flammability on a compound basis allows volatile emissions to be linked to both fire-related plant traits (cis-3-hexen-1-ol and myrcene with thermotolerance, β-pinene and myrcene with leaf water content) and climatic conditions (bornyl acetate, cis-3-hexen-1-ol and myrcene). This approach also provides useful information for the consideration of volatile plant compounds in future forestry practices and for the development of sensors to predict the onset of forest fires.