Morpho-Physiological Traits and Flammability of Bark in a Post-Fire Black Pine Population

Abstract

Pinus nigra Arnold, which is naturally widespread in mountainous and Mediterranean ecosystems, is a key species for reforestation due to its ecological and economic value. As climate change and changing fire regimes increase the wildfire risk, understanding its fire resilience has become critical. In this study, the morpho-physiological traits (thickness, roughness, moisture content) and flammability characteristics (ignition, heat release, mass loss, as determined in laboratory flammability tests) of the bark of P. nigra were investigated. The trees were selected based on their age (young vs. old) and fire exposure (burned vs. unburned). The bark thickness was significantly greater in older trees, while the bark moisture content was significantly lower in previously burned trees (p ≤ 0.05). The bark thickness correlated strongly with the ignition time, heat release, and mass loss. These results indicate that the age of the tree primarily affects the bark thickness and time to cambium death, while fire exposure primarily affects the bark moisture content, regardless of age. Understanding that the bark thickness and flammability play a key role in tree survival may aid in the selection of individuals or stand structures better suited to survive in fire-prone conditions and in the strategic planning of burns to reduce fuel loads without exceeding the mortality risk of younger or thinner-barked individuals.

1. Introduction

Wildfire is an ecological disturbance that alters the structure and composition of forest communities, affecting trees, understorey vegetation, and soil [1,2]. The fire resistance of forest communities refers to the ability of individual trees or forest stands to withstand fire without major damage, while resilience refers to the ability of a forest ecosystem to restore its structure and function after a fire event [3]. Understanding how certain forest types and species respond to fire is essential for the development of adaptive fire management strategies, especially in the face of climate change. Among conifers, pines (Pinus spp.) exhibit a wide range of fire-adaptive traits, with responses varying according to the species, stand structure, and local fire regime.

Pinus nigra J.F. Arnold is widely distributed in the mountainous regions of Europe and Asia Minor, but its populations are scattered and often patchily dispersed in a variety of ecological environments and climatic conditions [4]. In harsh environments, such as rocky soils exposed to strong winds and characterised by poor stand conditions, P. nigra tends to form pure and stable stands. The forest dynamics of P. nigra stands in mountainous regions with a Mediterranean climate (Spain, France and Turkey) are strongly influenced by their ability to persist under frequent fire regimes (with fire return intervals often of less than 10 years), but in regions with a more continental climate (south-eastern Alps and Dinaric Mountains), fire may not play such a major role and the forest dynamics are regulated by competition for light in sites with sufficient soil depth [4].

P. nigra is generally adapted to moderate fire frequency [3,5], but its survival is highly dependent on the fire type. Direct flame contact during surface fires can kill smaller trees, while medium-sized and larger individuals may survive, depending on their health and vigour. However, this survival is highly contingent on the fire type; black pine is particularly vulnerable to crown fires, which often result in high mortality across all size classes. Fire-caused injuries, whether from surface or crown fires, can predispose surviving trees to secondary pest attacks and fungal infections, contributing to post-fire decline and dieback [3]. As a member of the Diploxylon group (hard pines), P. nigra tends to have thicker bark and denser wood compared to Haploxylon species (soft pines), which generally have thinner bark and softer wood, traits that are often associated with lower fire resistance. In contrast to some pine species with serotinous cones (e.g., P. contorta), P. nigra generally regenerates from seeds released independently of fire, making it less specialised in fire-induced regeneration. Black pine has certain morphological traits, such as tall growth, self-pruning, and thick bark, that provide a certain resistance to low-intensity wildfires. However, its highly flammable needles and dense crown structure can contribute to increased fire intensity and facilitate the transition to crown fires, especially in dense, multi-layered stands. Thus, while black pine can tolerate surface fires under certain conditions, it remains highly susceptible to crown fires [6]. The species does not have a seed bank in the canopy, so the natural recovery of burnt stands after a fire depends on the dispersal of seeds from neighbouring, unburned areas or on surviving individuals of the species [7]. Studies indicate that the survival of P. nigra trees after fire is mainly dependent on the tree size [8], which has also been documented in other species [9,10]. The survival of larger trees is generally attributed to the thicker bark as the primary fire barrier that protects the cambium [11], as well as the higher crown base height, which allows a greater number of needles to be spared from fire [12]. Brando et al. [13] found that the tree mortality after a single, isolated fire in a transitional forest in the southern Amazon basin was relatively low. This suggests that forest resistance may be due to long-term adaptation to fire, with natural selection favouring fire-resistant individuals over time. However, they also pointed out that tree mortality is likely to increase with recurrent fires. Thus, the susceptibility of post-fire forest communities to repeated fire events remains ambiguous, as it reflects the interplay of two opposing processes: increased fire resistance due to the selective removal of fire-susceptible individuals and improved growing conditions for survivors, and increased fire sensitivity due to the increased susceptibility of fire-damaged trees to subsequent disturbances [13].

Natural and semi-natural P. nigra forests frequently exhibit uneven-aged stand structures due to historical practices, such as selective logging of dominant trees for economic yield, which have created openings in the canopy that promote natural regeneration, or natural disturbances, including windthrow and low-severity fires, which have played a pivotal role in creating gaps in the canopy that favour uneven-aged regeneration and multi-layered stand structures [4]. These stands are typically dominated by large, old trees, with several layers of younger, smaller trees and abundant deadwood from fallen individuals [14]. Forest stands with uneven-aged population structures are thought to be more resilient to certain natural disturbances, such as drought, windthrow, and insect or pathogen outbreaks, due to their structural diversity and regeneration potential (e.g., Mohr et al. [15]). However, in fire-prone ecosystems, these stands can be vulnerable to crown fires due to increased ladder fuel continuity unless the fuels are actively managed [16].

Fire management in P. nigra forests is increasingly based on integrated strategies aimed at reducing the risk of forest fires and improving the resilience of forests. In uneven-aged stands, which are typical of many natural and semi-natural P. nigra forests, mechanical thinning effectively reduces the canopy bulk density and vertical fuel continuity by selectively removing smaller, more flammable individuals that contribute to the formation of firewood ladders and facilitate the initiation of crown fires [17,18]. This approach favours the maintenance of dominant, thick-barked, fire-tolerant specimens that are more likely to survive low- to moderate-intensity fires and contribute to post-fire recovery [7,19,20,21]. However, the accumulation of logging residues after thinning can increase the aboveground fuel load and fire risk if not managed properly [22,23,24]. To mitigate this risk, prescribed burning is increasingly used in combination with thinning to reduce surface fuel, suppress understorey vegetation, and minimise fire severity. Fuel management strategies that combine thinning from below with low-intensity fire have been shown to be particularly effective in P. nigra stands [25,26,27]. In addition to reducing fuel loads, prescribed fires can eliminate the superficial fungal and litter crust around tree bases, which can inhibit seed germination, improving seed-to-soil contact and facilitating natural regeneration [28]. Recent studies have shown that although low-intensity fires can affect superficial fungal layers, they do not significantly alter the fungal community in the soil as a whole and thus maintain essential ecosystem functions [29]. In Mediterranean climates with prolonged dry summers, where the fire behaviour tends to be more extreme, the combination of thinning and prescribed burning has shown a synergistic effect in reducing the crown fire potential, flame length, and rate of spread compared to the application of either treatment alone [30,31].

Planned burning in uneven-aged stands poses a particular challenge, especially because of the younger trees that remain after thinning. These trees often have thin bark, which makes them much more susceptible to fire damage. Therefore, prescribed fire must be carefully adapted to the heterogeneous structure and fire sensitivity of such stands, to ensure that the fire intensity is low enough to avoid mortality in these susceptible cohorts [32,33,34]. Evidence from adaptive fire management programmes suggests that site-specific treatment combinations, which depend on the stand age structure, surface fuel load, and local climatic conditions, are essential for reducing the fire intensity while maintaining ecological integrity [35,36]. In southern Europe, where persistent fire suppression has led to excessive fuel accumulation in P. nigra forests, the application of integrated strategies combining mechanical and fire-based treatments is not only more effective but is also increasingly recognised as necessary to build long-term resilience in the face of increasing fire frequency and climate-related disturbances [30,37]. Detailed field data on the stand structure and key plant traits (bark thickness, canopy height, and regeneration patterns) are essential for the development of site-specific fire management strategies, as they help to identify individuals at risk and predict treatment outcomes. However, such data are often lacking at the local level, especially in P. nigra forests that have had little exposure to fire in the past and whose structural and adaptive responses to fire are poorly understood. This emphasises the need for baseline assessments to support effective, adaptive management in changing fire regimes.

It is well established that the bark thickness determines the persistence of the species and the ability to resprout [38,39]. There is often a negative correlation between bark thickness and resprouting ability: species that invest in thick bark are generally not resprouters, whereas species with thinner bark have a strong tendency to resprout after a fire [40]. However, this trade-off is not universal; it varies by species, fire regime, and plant functional type, and it is less pronounced in species adapted to low-intensity fire environments [41]. The survival of individual trees is largely determined by the length of the fire interval, which must be sufficient to allow the development of bark thickness adequate to resist subsequent fires [42]. In addition to its protective (insulating) function, the bark also serves as a store of non-structural carbon reserves, which are essential for recovery processes after a fire, such as resprouting, wound healing, and activation of the defences. However, these reserves can be severely depleted and/or not replenished under drought conditions. This is particularly evident in pines, where drought stress can significantly reduce carbon reserves and thus hinder post-fire recovery (see Piper and Paula [43]). The critical role of bark thickness in fire resistance has been extensively documented across different forest types, regions, and continents, and the fire resistance of tree species is often viewed through the lens of the bark thickness [44]. Recent studies on P. nigra confirm that the bark thickness is the dominant factor determining cambium heating and post-fire survival [45,46]. Bark thickness increases with tree size, which explains the higher survival rate of mature trees after fire, while young trees with thinner bark remain more vulnerable [47]. Under prescribed burning conditions, trees with sufficient bark thickness generally exhibit low cambium heating, minimal physiological impairment, and negligible growth reductions [48,49], supporting the view that thick bark is central to P. nigra’s persistence under infrequent, low-intensity fire regimes [50]. Only a few studies have investigated the relationships between the bark traits, bark flammability, and fire resistance of P. nigra, and most of them have been conducted in the Western Mediterranean Basin. To the best of the authors’ knowledge, there are no published studies addressing these relationships in black pine populations in the Balkan Peninsula, despite the ecological and economic importance of the species in this region. This knowledge gap limits our understanding of how local environmental conditions and fire regimes influence bark properties and fire tolerance in these forests. The aim of this study was therefore to investigate the flammability and the morpho-physiological traits of the bark of P. nigra in a stand affected by a surface fire, with the specific objectives being to: (i) assess the variability of bark traits in relation to tree age and fire exposure, and (ii) examine the relationships between bark flammability and morpho-physiological characteristics.

2. Materials and Methods

2.1. Study Site

The study site is located on the Murtenica mountain at 43°34′17″ N, 19°47′11″ E, on a moderate slope (11–15%) facing south–southwest (the location map is shown in Figure 1a). The elevation ranges from 1230 to 1431 m, and the substrate is brown forest soil developed on limestone. The forest comprises a pure black pine stand with a dense assemblage (density level = 0.8–0.9, referring to canopy closure) and good health. Field observations and qualitative assessments indicate an uneven age structure, characterised by a range of tree sizes and poor regeneration. This is confirmed by the latest forest inventory data [51] and illustrated in the diameter distribution graph (Figure 1b). The understorey vegetation is dense, consisting primarily of Festuca drymeia, Pteridium aquilinum, Corylus avelana, and Juniperus communis, with abundant dead cover in a thin layer. A surface fire burned approximately 20 hectares of forest in August 2021. A vegetation survey was conducted in August–September 2024, three years after the fire, which is the minimum time frame for assessing delayed mortality [52].

2.2. Plant Material

Within the stand, there were trees affected by fire, with charred bark, and those without visible signs of burning. The burn status was determined visually based on the presence of superficial charring on the bark and burn marks on the ground (e.g., ground char). Trees were classified as “burned” (B) if the charring extended deeper than 0.5 cm into the bark and was not just superficial. “Unburned” (UB) trees showed no visible charring, either on the surface or in depth. The age of the tree was quantified by extracting increment cores at the DBH (diameter breast height, 1.35 m above the ground) and counting the annual rings. On this basis, the specimens were classified into two groups: young (Y, 15–20 years) and old (O, 40–60 years). Sampling followed a systematic spatial design along a strip line from the forest edge to approximately 10 m inside the stand. Five to seven trees from each group were selected for bark sampling: young burned (YB), old burned (OB), young unburned (YUB), and old unburned (OUB). To reduce sampling bias, neighbouring trees were avoided, and a minimum distance of 10 m was maintained between sampled individuals. The selected trees were in a good state of vitality, with approximately 80% green needles and without visible damages caused by insects or pests. On burned trees, char was visible on the bark at a height of 0.4 m (downhill side of trees) to 0.75 m (uphill side of trees).

The bark samples were taken at a height of 0.45 to 0.70 m, which corresponds to the upper limit of bark charring on burnt trees and covers ecologically relevant fire exposure zones typical of low-intensity surface fires [53]. The entire bark was removed with a sharp wood-carving knife up to the visibly distinct cambium interface (identified by its smooth texture and lighter colour beneath the inner bark), packed in airtight plastic bags and immediately transported to the laboratory. Bark pieces were collected from different stem regions (uphill and downhill sides of the tree) to cover inter-individual variability in moisture, bark thickness and surface structure. From each tree, the bark pieces were sorted into sub-samples with approximate dimensions 7 × 7 cm, placed in plastic bags, and sealed shut for transport to the laboratory. Six sub-samples were used for determination of morphological traits (thickness, roughness) and six sub-samples were used for determination of the moisture content and burn testing. A flowchart illustrating the sampling procedure is presented in Figure 2.

2.3. Bark Morpho-Physiological Traits

The moisture content (MC, %) of six sub-samples from each tree was determined gravimetrically by weighing fresh samples (F) and reweighing after 48 h of drying at 105 °C (D). The moisture content was calculated on a wet basis (percentage of fresh weight) as follows: MC = (F − D)/F, where F and D represent the weight of fresh and dry samples, respectively. The bark thickness and bark roughness were determined on six bark pieces (sub-samples) from each tree with approximate dimensions of 7 × 7 cm. The bark thickness (BT, cm) was measured with a digital micrometer (Käfer Messuhrenfabrik GmbH & Co., Villingen-Schwenningen, Germany) in four positions in each sub-sample (maximum and minimum bark thickness, and two additional random points) on the most representative faces (n = 6 × 4 = 24 per tree), and the bark thickness per individual tree (BT) was calculated as an average of four measurements. The degree of bark roughness (BR) for a given BT was calculated as following: BR = max (BT) − min (BT)/max (BT).

The bark insulating ability was determined by the critical time to cambium kill (τc, min), which was calculated according to the formula reported by Peterson and Ryan [54]: τc = 2.9 × BT², where τc is the critical time for cambium kill and BT is the bark thickness.

2.4. Bark Flammability

Burning experiments were conducted at the Fire Protection Laboratory, Faculty of Occupational Safety, University of Niš, Serbia. The tests were performed using a mass loss calorimeter, a device that enables controlled exposure of samples to a defined heat flux while simultaneously monitoring the mass loss, heat release rate, and other combustion parameters [55]. For the purpose of this study, the cone heater of the mass loss calorimeter was calibrated to a heat flux value of 50 kW/m². At this heat flux value, the heater temperature reached 778 °C, which is in good agreement with the data presented by Babrauskas [56]. The operation and calibration of the mass loss calorimeter are described in detail in the Instruction Manual [57]. A custom-made sample holder measuring 10 × 10 × 3.5 cm with perforated sides was used. Figure 3 shows a representative bark sample placed in the sample holder before the experiment, as well as the combustion process during the experiment. The perforation on the sidewalls accounted for approximately 63% of the total surface area of the sample holder. A sample holder of a similar design was used in the same manner as in the study by Schemel et al. [58], in which it was shown that perforation allows for controlled transport of gases through the porous structure of the fuel. This design of the sample holder allows for adequate air circulation through the sample during combustion and more realistically simulates natural burning conditions during forest fires. The distance between the sample holder and the edge of the cone heater was 25 mm. The mass of each sample was approximately 20 g ± 0.5 g. In all the experiments, piloted ignition was applied, positioned at the centre of the sample to ensure faster ignition of the volatile gases generated after exposure to thermal radiation. Eight components of flammability were assessed: time to ignition (TTI, s), peak heat release rate (Peak HRR, kW/m²), peak effective heat of combustion (Peak EHC, MJ/kg), peak mass loss rate (Peak MLR, g/s), mean heat release rate (Mean HRR, kW/m²), mean effective heat of combustion (Mean EHC, MJ/kg), mean mass loss rate (mean MLR, g/s), and total heat release (THR, MJ/m²). The mass loss rate was monitored using a measuring cell with an accuracy of 0.1 g and a response time of 90% in less than three seconds. The bark flammability parameters were quantified for trees classified in four groups (YB, OB, YUB, OUB). During the experiments, all the samples successfully ignited. Throughout the experiments, the laboratory temperature was 28 ± 2 °C, with a relative humidity of 57 ± 3%. The minor fluctuations did not affect the obtained experimental results.

2.5. Statistical Analyses

To avoid pseudoreplication bias, multiple measurements per tree were averaged before the statistical analyses. All the analyses were performed using R statistical software (version 4.2.1) [59], employing various statistical methods. ANOVA and the Kruskal–Wallis rank sum test with the false discovery rate (FDR) p-adjustment method were used to assess the differences between selected groups of P. nigra in terms of the morpho-physiological bark traits (BT, BR, MC, τc) and flammability parameters (TTI, Peak HRR, Peak EHC, Peak MLR, Mean HRR, Mean EHC, Mean MLR, THR). The normality of the distribution was tested using the Shapiro–Wilk test. The significance level (α) was set at 95% for all the statistical tests (p ≤ 0.05). Principal component analysis (PCA) was applied to all the measured variables to assess whether there was a grouping of individuals. Correlation analysis was performed to explore the relationships between traits within a single biological population of P. nigra. Although the dataset includes individuals from four groups that differ in age and fire exposure, all the trees belong to the same population and were analysed together to identify general patterns of association, not group-level effects.

3. Results

Young and old burned and unburned individuals of P. nigra differed in their bark morpho-physiological and flammability traits (Figure 4). Compared to old trees, young trees had significantly lower values for bark thickness and critical time to cambium kill, and burned trees had significantly lower bark moisture content compared to unburned trees. Young burned trees had the highest values for bark roughness. Old unburned trees took the longest and young burned trees the shortest time to ignite. The EHC values showed certain differences between the groups. The EHC peak and mean values in the OB samples were higher than the values in YB, YUB, and OUB. This indicates that the bark in OB is capable of releasing a greater amount of energy per unit mass of sample at the time of the most intense combustion.

The combustion process can be divided into three phases: the preheating phase, the flaming phase, and the smouldering phase. Immediately after positioning the sample under the heating cone, the preheating phase begins, during which volatile compounds are released. When the concentration of volatile substances reaches the lower flammability limit, piloted ignition occurs, leading to the appearance of flames that soon spread over the entire sample surface. Figure 5 shows a representative sample together with the curves of the development of the flammability parameters: HRR, MLR, THR, and EHC. Immediately after ignition, the HRR rises sharply and quickly reaches its peak value. Thereafter, the HRR gradually decreases while maintaining relatively high values due to the continued smouldering of the charred sample bark. Simultaneously with the development of the HRR, the MLR increases rapidly in the early combustion phase, immediately after ignition, when most of the volatile compounds are released from the samples. After reaching the peak value, the MLR gradually decreases, corresponding to the reduction of the remaining combustible fraction in the samples. Throughout the experiment, the THR increases continuously as the combustion process progresses. The EHC initially shows low values in the early combustion phase, but as the flaming phase develops and the combustion process stabilises, its value increases.

The principal component analysis resulted in a three-component model, with the first three PC axes explaining 71% of the variability in the data (Figure 6). In the plane of the first two PC axes, a separation of the young individuals (burned and unburned) can be seen, while the old individuals overlap (Figure 6a). In the plane of the first and third PC axes, a better separation of OB and OUB individuals can be seen (Figure 6b). The BT, τc, Peak MLR, Mean EHC, and TTI had the greatest influence on the formation of PC1 and on the separation of young (YB, YUB) and old individuals (OB, OUB); higher values for these variables were found in old individuals. The Mean MLR, Mean HRR, MC, and Peak HRR had the greatest influence on the formation of PC2 and on the separation of young burned and unburned groups; YB trees had lower values for the Mean MLR and MC. The formation of PC3 was predominantly influenced by the THR, Mean EHC, TTI, MC, and Mean HRR, and these variables influenced the separation of old burned and unburned trees; OB trees were characterised by higher Mean EHC and lower TTI and MC compared to OUB trees.

The relationships between the measured variables were quantified using the Spearman’s correlations (

Table 1. Spearman’s correlation coefficients between the measured variables. Statistically significant correlations are shown in bold. Asterisks indicate statistical significance: * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001.

	BT	BR	TC	MC	TTI	Peak HRR	Peak EHC	Peak MLR	Mean HRR	Mean EHC	Mean MLR	THR
BT	1
BR	−0.32	1
TC	1 ***	−0.32	1
MC	−0.04	−0.09	−0.04	1
TTI	0.52 **	−0.21	0.52 **	0.2	1
Peak HRR	0.08	−0.12	0.08	0.34	−0.19	1
Peak EHC	0.35	−0.18	0.35	−0.41 *	0.12	−0.14	1
Peak MLR	0.65 ***	−0.34	0.65 ***	−0.01	0.64 ***	−0.13	0.45 *	1
Mean HRR	−0.09	−0.14	−0.09	0.26	−0.16	0.61 **	−0.2	−0.08	1
Mean EHC	0.48 *	−0.4	0.48 *	−0.26	0.22	0.12	0.34	0.53 **	0.47 *	1
Mean MLR	0.13	−0.24	0.13	0.59 **	0.09	0.29	−0.28	0.18	0.39	0.05	1
THR	0.06	−0.14	0.06	−0.41 *	−0.34	0.23	0.11	0	0.52 **	0.74 ***	−0.15	1

). The BT and τc were positively correlated with the TTI, Peak MLR, and Mean EHC, while the MC was positively correlated with the Mean MLR and negatively correlated with the Peak EHC and THR.

4. Discussion

Bark’s morpho-physiological traits and flammability components contributed to the discrimination of black pine individuals by (i) age class (PC1 axis), with the BT, τc, Peak MLR, Mean EHC, and TTI being the traits with the largest contribution, and (ii) burned vs. unburned status (PC2 and PC3 axes), with the Mean MLR, Mean HRR, MC, and PHRR mainly contributing to the discrimination between young individuals and the THR, Mean EHC, TTI, MC, and Mean HRR mainly contributing to the discrimination between mature individuals. The stronger separation was observed based on age-defined groups, and mature individuals were characterised by thicker bark, longer time to cambium death, longer time to bark ignition, and higher values of specific combustion characteristics (Mean EHC, Peak EHC). Previous studies reported that the thermal insulation capacity of bark is determined by the bark thickness [13,42,60,61], which was confirmed for mature individuals. Based on the BT and τc, used in this study to assess the insulating capacity of the bark, which were positively correlated with the TTI, younger individuals were at higher risk of rapid ignition than mature individuals. Bark thickness is generally considered a critical feature, which provides sufficient insulation to protect the cambium from lethal temperatures during fire exposure, and studies report that a bark thickness of ≤17 mm may be the threshold for cambium death in P. nigra [45]. However, the variability of the bark thickness, which is often higher in older specimens, may also play a role leading to localised damage to cambium tissue, even if the average bark thickness is adequate [40]. Finally, high-intensity and prolonged exposure increase the risk of cambium damage, regardless of bark thickness. The shorter time required to reach a lethal cambium temperature (which correlates strongly with the bark thickness) is a critical parameter for assessing resistance to tissue damage in low- to moderate-intensity fires. The density of the bark also contributes to heat transfer, and bark with low density and more air-filled interstices can result in lower thermal conductivity even with the same bark thickness [62,63,64]. It should also be noted that the separation of outer and inner bark to study the bark’s morpho-physiological traits and flammability would allow deeper insights into the protective function of the bark, given the different structural and functional characteristics of these layers.

Using a series of flammability tests on P. nigra bark, conducted with a cone calorimeter at a heat flux of 20 to 70 kW/m², a range that covers both prescribed burning and wildfire conditions, Espinosa et al. [45] determined a bark thickness threshold of 17 mm, at which the probability of cambium death decreases. The average bark thickness of all the samples in our study was between 4 and 14 mm, which is below the reported thresholds and indicates that the cambium was probably partially damaged in the charred zone of the bark, especially in young specimens. Studies involving heat girdling of the stem found that nearly 100% cambium injury is required to kill a tree [65,66], and in the absence of significant crown injury, most trees survive up to 25% basal girdling (cambium kill), but few trees survive more than 75% girdling [67]. This may explain the survival of the trees selected in this study, although the bark thickness was below the threshold for cambium kill. The results showed that burned P. nigra specimens, especially YB, exhibited significantly higher bark roughness compared to unburned specimens. In fire-prone ecosystems, bark roughness, especially deep fissures, can increase fire resistance by creating insulating air pockets that protect the cambium. In a study by Shearman and Varner [68], fire-adapted tree species were shown to have protective outer bark near the base of the bole and variable bark roughness that can buffer the effects of surface fire. The highest value observed in the YB group may indicate an early investment in bark, meaning that trees disproportionately invest biomass in bark early in their lives, even before they reach large diameters, to enhance survival in stressful environments such as habitats with frequent fires [69].

While the BT and τc were strongly determined by age class, the MC was significantly lower in burned individuals than in unburned individuals, regardless of age. The lower MC in B individuals compared to UB individuals may be the result of heat exposure, vascular tissue damage and post-fire stress physiology (reduced transpiration and impaired water uptake due to root damage). Following a fire, trees often show reduced transpiration and photosynthetic activity as a defence mechanism to conserve water and energy when vascular function is impaired or foliage is partially lost [70,71]. These physiological adaptations can further reduce the water content in the bark and cambium, especially in trees that are partially scorched but survive the fire. Burned individuals exposed to surface fire may suffer partial root system damage, either through soil heating or indirectly through altered soil microbial communities, that may reduce the water uptake capacity [72]. Structural changes in the bark, reflected in the increased roughness of burned specimens, can increase the surface area and moisture loss. In addition, post-fire physiological stress, including carbon depletion and hormonal imbalance, can disrupt stomatal regulation and water balance, resulting in increased water loss or limited water retention [73]. The MC is generally considered the most important plant trait affecting the flammability of various living and dead plant organs and is strongly correlated with ignition, as it delays ignition since much of the heat is absorbed by heating and vaporising the water [74]. The MC was not significantly correlated with the TTI in this study but was negatively correlated with other combustion components (Peak EHC and THR). Studies have shown that increased moisture content in plant material significantly reduces the peak heat release rate [75]. The thick bark with high water content could be the most favourable trait for the insulating function. However, Poorter et al. [76] found a strong negative correlation between the density and water content of bark and wood in 50 woody species from dry and moist forests, suggesting that either solid structural material or water is allocated to the bark, while gas contributes only a small part.

The structural complexity and diversity of uneven-aged forests influences fire behaviour. The presence of small diameter trees with dense regeneration can increase the crowning fire risk, supporting thinning from below as an important mechanical measure to reduce this risk [23,77]. Depending on the extent of the fire injury and pre-fire condition, surviving trees have different recovery rates and capacities, and the performance of individual trees and the whole stand depends on the trade-off between the different traits. Younger trees tend to recover faster due to higher growth rates and intense physiological processes but are also more susceptible to fire damage because their bark is thinner and they are exposed to prolonged heating from surface fuels due to their proximity to the ground [42]. Older and larger trees, which have thicker bark, can shield vital tissues from the heat but may face other limitations to rapid and full recovery, such as senescence, pathogen loads, and reduced energy reserves [78]. Functional recovery from cambium injury requires sufficient carbon reserves and efficient resource mobilisation, which can only be achieved if the tree maintains high vigour [48,70].

On the basis of bark traits (bark thickness, time to cambium kill and normalised bark thickness, i.e., the ratio of bark thickness to stem radius), Fernandes et al. [79] ranked P. nigra between montane pines such as P. sylvestris and P. uncinata and lower-altitude Mediterranean pines such as P. pinaster, P. pinea, and P. halepensis. However, empirical data establishing a correlation between bark thickness (and other bark traits) and flammability are scarce. Such correlations are crucial for understanding the bark-related fire adaptation strategies of pines. Comparative studies would be particularly valuable at sites where different coexisting species experience similar climatic and environmental conditions, as fire adaptations may be better analysed at the population level than at the species level. For example, P. halepensis coexists with P. nigra in various mixed communities in the Iberian Peninsula, southwestern Anatolia, the eastern Mediterranean, and the Balkans [80]. A comprehensive study of 19 provenances of P. halepensis found that the total bark thickness (including inner and outer layers) was greater in populations from areas with frequent fire activity (23.5 mm and 12.5 mm at 10 cm and 130 cm above the ground, respectively). In contrast, populations from regions with high drought and short growing seasons had thinner bark (20.6 mm and 8.1 mm at the same tree heights, respectively). These results support the ecotypic nature of bark traits and clear the way for further research on critical thresholds for bark thickness that are crucial for fire survival [81]. Young P. pinaster trees are reported to be less fire-resistant, and prescribed fire is generally avoided when the DBH is below 10 cm, with mechanical clearing recommended as the preferred treatment [82]. In another study with P. sylvestris and P. nigra, which often form mixed forests in central and southern Europe [80], it was found that the thickness of the bark after a fire and the area of the resin canals are important factors influencing the susceptibility of the trees to bark beetles. The authors suggested that the lower post-fire mortality observed in P. nigra compared to P. sylvestris could be due to a smaller reduction in the bark thickness with increasing stem height in P. nigra, even though the stem diameter increases [83]. Bark traits not only affect the resistance of the trees but also influence the flammability of the litter. Bark fragments contribute significantly to the surface fuel and influence the ignition and spread of fire. Studies have shown that thicker or more complex bark fragments can form a denser litter bed that retains moisture and burns less intensely [84,85,86].

Black pine forests are increasingly threatened by fires, not only in traditionally fire-prone Mediterranean and sub-Mediterranean regions but also in historically less fire-prone temperate or montane zones due to the increasing effects of climate change [87]. Mount Murtenica is located in the Inner Dinarides region, with a moderate continental and mountain climate, which is expected to change due to rising temperatures, prolonged droughts, and lower precipitation [88], potentially leading to weather conditions that favour frequent fires and lower the ignition thresholds for vegetation. As previous research on bark traits and flammability has focused on mature trees, this study provides information on age-related differences that are important for fire management in uneven-aged forest communities. In particular, the data suggesting that the critical time for cambium death is about five times shorter in young trees than in mature trees, regardless of whether they have been previously exposed to fire, may help to balance the intensity and duration of fire treatments. The heat transfer and temperature distribution within the bark were beyond the scope of this study. However, as these factors are important for predicting tree survival during wildfires, future research will incorporate thermocouples on bark samples to record these measurements during exposure to heat fluxes. A deeper understanding of the stand structure, tree age, bark traits, and flammability can improve the assessment of individual tree survival, especially in low- to moderate-intensity fires, as well as fire prevention through controlled fire treatments designed to ensure the survival of the most vulnerable age classes.

The results presented herein provide complementary insights to existing findings from Mediterranean pine ecosystems, where low fuel moisture and active crown fires often lead to severe ecological impacts [89,90]. While no silvicultural treatments or fire suppression measures were directly evaluated in this study, the observed relationships between bark traits and flammability highlight the need to integrate stand structure into fire prevention strategies. The key indicators for fire management include the (i) bark thickness thresholds (protecting younger cohorts during burns); (ii) time to cambium death (as young trees show about five times shorter τc than mature trees, controlled burns should be designed to limit fire exposure); (iii) moisture content (lower in burned trees, indicating higher flammability and post-fire stress; monitoring MC pre- and post-fire can inform on tree stress and susceptibility, guiding timing of burns to periods of higher moisture to reduce ignition risk); and (iv) fuel structure (thinning from below can reduce ladder fuels and protect vulnerable cohorts when used in conjunction with prescribed fires).

5. Conclusions

Climate change is contributing to fire-prone conditions, increasing fuel accumulation, and lowering ignition thresholds even in traditionally non-fire-prone areas, necessitating improved risk mitigation strategies.

The vulnerability of black pine stands is strongly influenced by the stand structure, tree age, and functional traits, with bark thickness emerging as the most important protective trait against fire.

This study showed that the morpho-physiological traits and the flammability of the bark of P. nigra depend significantly on the age class and previous fire exposure.

Older trees exhibited greater fire resistance, characterised by thicker bark, delayed ignition, and higher combustion efficiency, while younger trees were more susceptible due to their thinner bark and faster ignition.

Although the average bark thickness was below the reported threshold for complete cambium protection, exposure to fire did not result in widespread mortality, suggesting that even sub-threshold bark thickness can provide protection during low-intensity surface fires.

The key indicators for effective fire management include the bark thickness, time to cambium death, bark moisture content, and overall fuel structure. Integrative strategies, such as thinning from below to reduce the risk of crown fire, protecting sensitive age groups during prescribed fires, and carefully timing burns based on monitoring the bark moisture content, can improve resilience and minimise fire-related mortality.

These results are also relevant for other fire-prone ecosystems with uneven-aged structures, where the inclusion of tree traits in fire management can improve the effectiveness and ecological sensitivity of burning measures.