Non-Destructive Methods for Diagnosing Surface-Fire-Damaged Pinus densiflora and Quercus variabilis

Abstract

Wildfires impact forest ecosystems, affecting tree survival and physiological responses. This study explored the effects of surface fires on Pinus densiflora and Quercus variabilis, assessing mortality, internal injuries, and canopy health. By 2024, P. densiflora had an 18.0% mortality rate, whereas Q. variabilis exhibited no crown dieback. Morphological traits, including tree height, the bark scorch index (BSI), and bark thickness, influenced fire resistance. Despite superior stand characteristics, P. densiflora showed higher mortality due to thin bark, whereas Q. variabilis maintained xylem integrity. While sonic tomography (SoT) showed no significant differences, electrical resistance tomography (ERT) detected physiological stress, with higher ERT_R and ERT_Y area ratios correlating with mortality risk. Notably, F-W-W classified trees showed elevated resistance a year before mortality, suggesting ERT as a predictive tool. ERT_R values exceeding 15.0% were associated with a 37.5% mortality rate, whereas ERT_B values below 55.0% corresponded to 42.9% mortality. Despite fire exposure, canopy responses, including chlorophyll fluorescence and photosynthetic efficiency, remained stable, indicating that the surviving trees maintained functional integrity. This study underscores ERT’s efficacy in diagnosing fire-induced stress and predicting mortality risk. The findings highlight species-specific diagnostic criteria and inform post-fire management, supporting forest resilience through the early detection of high-risk trees and improved restoration strategies.

1. Introduction

Between 2014 and 2023, the annual average number of wildfires in the Republic of Korea was reported to be 567, with an affected area of 4004 ha. Notably, in 2022, large-scale wildfires exceeding 100 ha occurred 11 times in regions such as Uljin–Samcheok and Gangneung–Donghae, resulting in damage to approximately 24,787.5 ha of forest [1]. In the Republic of Korea, recent large-scale wildfires have been associated with unusually prolonged periods of low humidity and high temperatures, particularly during the spring fire season. According to the Korea Meteorological Administration, these dry atmospheric conditions have become increasingly frequent and intense, a trend attributed to broader climate change effects including rising temperatures and shifting precipitation patterns [2,3,4,5,6]. Empirical observations in Korean forest regions show that warming trends and reduced spring rainfall have contributed to drier forest floor conditions, which, in turn, increase the probability and severity of wildfire outbreaks [7,8].

Wildfires are generally classified as ground fires, surface fires, or crown fires. Ground fires burn organic matter beneath the forest floor, while surface fires spread through low vegetation such as grasses and shrubs. Crown fires, the most intense type, consume the canopy layer and often result in severe ecological damage. Among these, surface fires are characterized by relatively low intensity and a short duration but occur frequently, posing a persistent threat to forest ecosystems [9]. The impacts of wildfires extend beyond direct forest damage, leading to biodiversity loss, soil degradation, atmospheric changes, and disruptions in aquatic ecosystems. In wildfire-affected areas, the absence of standing trees results in drier and more compacted soil, accelerating nutrient leaching and topsoil erosion, which ultimately contribute to land degradation. Additionally, large-scale sediment runoff can weaken forest lands, increasing the likelihood of secondary disasters such as landslides [10,11,12]. Furthermore, particulate matter generated by wildfires may be transported to coastal ecosystems via river systems, causing further environmental disturbances [13].

When wildfire-induced soil temperatures rise to approximately 48–54 °C, soil moisture evaporation may lead to root desiccation, which can contribute to tree mortality depending on species-specific heat tolerance and environmental conditions [14,15]. For example, Korean native species such as Pinus densiflora and Quercus variabilis may exhibit varying physiological thresholds, and further research is needed to establish precise critical limits. The degree of damage also depends on fire intensity and duration [16,17], and in most surface fires, the insulating properties of soil reduce the extent of heat damage to roots [18]. Tree mortality following wildfire events is driven by a combination of factors, among which direct heat-induced tissue necrosis is often the most immediate and severe. Other contributing mechanisms include the disruption of water and nutrient transport, imbalances in photosynthesis and respiration, and increased susceptibility to pathogens and pests [19,20,21]. Fire temperatures vary considerably with intensity and fuel load. Under extreme fire conditions—such as those associated with high fuel availability and windy, dry weather—temperatures may exceed 650 °C at the flame front, while moderate and low-intensity surface fires typically reach around 400 °C and 250 °C, respectively [22]. Even at lower temperatures, prolonged exposure can cause irreversible cellular damage in plant tissues [23,24].

Tree species exhibit varying degrees of fire resistance, with broadleaf species generally being more fire-resistant than conifers [25]. In the Republic of Korea, which falls within the temperate climate zone, Quercus species such as Quercus acutissima and Quercus variabilis dominate in terms of distribution, while P. densiflora represents the most widely distributed single species, making it a representative species of the region. Q. variabilis is known as a highly fire-resistant species due to its thick cork layer, which has low thermal conductivity and short flame persistence [26]. In contrast, P. densiflora has a lower crown height and thinner bark, making it more susceptible to horizontal fire spread as well as vertical fire progression from surface fires to trunk and crown fires [27]. Due to these characteristics, the bark scorch index (BSI) was developed as a rapid field assessment tool to predict the survival of P. densiflora affected by surface fires [28].

Although trees affected by surface fires may sustain varying degrees of damage, some individuals have been reported to survive and recover without requiring artificial reforestation under moderate fire conditions, typically characterized by flame temperatures around 250–400 °C and low to moderate surface fuel loads [29]. In such cases, maintaining the existing forest structure offers ecological and economic advantages. Therefore, it is crucial to develop non-destructive methods for diagnosing surface fire damage and assessing the retention potential of affected trees. To address this need, the present study employs trunk-based non-destructive diagnostic techniques—sonic tomography (SoT) and electrical resistance tomography (ERT)—to evaluate internal damage patterns in fire-affected trees. In the Republic of Korea, these methods have previously been applied to assess internal decay in large, old trees and protected trees [30,31,32], and to diagnose fungal decay [33,34]. ERT has also been used to examine sapwood–heartwood compartmentalization in various species [35,36,37]. However, applications for post-fire diagnosis remain limited and largely exploratory, with this study contributing new insights based on assessments of P. densiflora and Q. variabilis.

This study aims to evaluate internal damage and mortality in P. densiflora and Q. variabilis trees affected by surface fires using SoT, ERT, and morphological indicators. The following research questions are addressed: (1) What are the characteristic internal damage patterns of each species following surface fire exposure? (2) How do morphological traits correlate with internal damage severity and tree survival? (3) What are the key species-specific factors that influence fire resistance? The findings are expected to inform species-level fire vulnerability assessments and enhance post-fire forest management and decision-making regarding tree retention and natural regeneration potential.

2. Materials and Methods

2.1. Study Site

This study was conducted in Cheonpo, Sindong, Jeongseon, Gangwon State, Republic of Korea (37° 11.777′ N, 128° 38.378′ E), where a wildfire occurred on 29 October 2020 (Figure 1). The wildfire was classified as a surface fire with a low intensity level, affecting approximately 0.4 ha of forest. The study site is a mixed forest dominated by Pinus densiflora and Quercus variabilis, situated at an elevation of 520–565 m, with a mean slope of 28.4 ± 5.1° on a southwest-facing slope.

In April 2022, trees that remained standing after the wildfire were selected, considering differences in species resistance. Totals of 50 P. densiflora (Pd) and 30 Q. variabilis (Qv) individuals were chosen. Additionally, 10 Pd and 5 Qv trees with no visible scorch marks on the bark, soil, or understory vegetation were selected as non-fire-damaged (NFD) controls. Subsequent investigations in 2023 and 2024 were conducted during spring (April–May) to minimize seasonal variations.

2.2. External Morphological Characteristics

For each selected tree, the tree height (TH) and bole length (BL) were measured using a clinometer, while the diameter at breast height (DBH) and circumference at the measuring point (CircMP) were recorded using a diameter tape. Canopy condition was visually assessed and classified as either fine (leaves present and green) or withered (completely defoliated or brown) (Figure 1).

The bark scorch index (BSI) was assessed by dividing the trunk at breast height into four quadrants and measuring two parameters in each: the bark scorch height (BSH, the vertical height of visible bark scorch from the base) and bark scorch proportion (BSP, the percentage of visible bark scorch from the base). The BSI was then calculated as the sum of the products of BSH and BSP across all quadrants [28] (Figure 1). Bark thickness was measured at MP height using a steel ruler, recording both the thickest and thinnest points to calculate the average.

2.3. Trunk Internal Damage Diagnosis

Internal trunk assessments were performed using geometric tomography (Geo), sonic tomography (SoT), and electrical resistance tomography (ERT). Measurements were taken at 0.4 m above ground level, a height selected based on preliminary field surveys, to minimize interference from understory vegetation (generally < 30 cm in height) and to ensure consistent data acquisition across trees with varied scorch distributions. The measuring points (MPs) were standardized to eight per tree, maintaining minimum spacing to reduce mechanical error. The northernmost MP was consistently designated as point 1 for comparative analysis.

Geometric tomography was conducted using a PiCUS Calliper Version 3 (Argus Electronic, GmbH, Rostock, Germany) to measure MP distances, converting data into geometric models compatible with SoT and ERT. SoT was performed using PiCUS 3 Sonic Tomography (Argus Electronic, GmbH, Rostock, Germany) with sensors attached at MPs and struck using a PiCUS electronic hammer (Argus Electronic, GmbH, Rostock, Germany) [38]. Image quality was refined through cogwheel removal [39].

ERT was conducted at the same MPs using a PiCUS Treetronic device (Argus Electronic, GmbH, Rostock, Germany) [40]. Image enhancement involved adjustments in scale, smoothness, and mesh fineness, extracting the color area ratio (CAR) and resistivity (Rs) to assess fire-induced damage. Since fire damage results in localized high resistivity, the minimum (Rs_Min) and maximum resistivity (Rs_Max) values were recorded. Additionally, considering that water transport occurs primarily in sapwood, we applied the criterion t/R < 0.3 (t = residual wall thickness; R = tree radius) to distinguish sapwood for further analysis [39]. Abbreviations used in this study are summarized in

Table 1. Abbreviated terms used in the experiment.

Abbreviations	Descriptions
General terms
P. densiflora; Pd	Pinus densiflora
Q. variabilis; Qv	Quercus variabilis
FD	Fire-damaged
NFD	Non-fire-damaged
Terms related to external appearances
TH	Tree height
BL	Bole length
DBH	Diameter at breast height
CircMP	Circumstances at measuring point
BSI	Bark scorch index
BSH	Bark scorch height
BSP	Bark scorch proportion
F-F-F	Fine from 2022 to 2024
F-W-W	Fine until 2022 and withered from 2023
W-W-W	Withered from 2022
Terms related to internal diagnosis
Geo	Geometry
SoT	Sonic tomography
ERT	Electrical resistance tomography
CAR	Color area ratio
RsMin	Minimum resistivity
RsMax	Maximum resistivity

.

2.4. Canopy Physiological Responses

Physiological responses were assessed through the pigment content, photosynthetic and stomatal responses, and chlorophyll a fluorescence analysis. The study was conducted in the period of April–May 2022, approximately one year and a half post-fire. Only individuals that retained visually green leaves and were not completely withered were selected. Seven fire-damaged and three non-fire-damaged trees per species were chosen. Branch samples were collected from the upper canopy, ensuring full sun exposure and accessibility within an 8–10 m height range. Samples were at least 30 cm long and were immediately placed in water after cutting to preserve physiological conditions. The entire sample was wrapped in black plastic to maintain moisture and reduce transpiration [41,42].

Pigment content analysis followed [43], using 0.1 g of leaf tissue extracted in 10 mL of dimethyl sulfoxide (DMSO) at 65 °C for 6 h. Absorbance was measured at 663 nm, 645 nm, and 445 nm using a Multiskan Go Microplate Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Chlorophyll a and b, total chlorophyll, and carotenoid contents were calculated, along with chlorophyll a/b and total Chl/Car ratios [44,45].

Chlorophyll a fluorescence was measured using a Plant Efficiency Analyzer (PEA, Hansatech Instrument Ltd., King’s Lynn, UK). Leaves were dark-adapted for 20 min before exposure to 3500 µmol·m⁻²·s⁻¹ light for 1 s, and fluorescence intensities were recorded at O (50 µs), K (300 µs), J (2 ms), I (30 ms), and P (500 ms) phases. Biophysical parameters were derived from OKJIP analysis [46,47] (

Table 2. Descriptions of abbreviated parameters related to OJIP test.

Abbreviations	Descriptions
ABS/RC	Absorption flux per RC
DIo/RC	Energy dissipation flux per RC
TRo/RC	Trapped energy flux per RC
VK/VJ	Ratio of variable fluorescence in K to J step as indicator of PSII donor side limitation
ϕPo	Maximum quantum yield of primary photochemistry
ψEo	Efficiency with which trapped exciton moves electron into ETC beyond QA
ϕEo	Quantum yield of electron transport
PIABS	Performance index on absorption basis
DFABS	Driving force on absorption basis
SFIABS	Structural and functional index on absorption basis

).

Photosynthetic and stomatal responses were assessed using a portable photosynthesis system (Li-6800, Li-Cor Inc., Lincoln, NE, USA) by measuring the maximum photosynthetic rate (A_Max), stomatal transpiration rate (E), stomatal conductance (gs), and intercellular CO₂ concentration (Ci). Measurements were taken under a photosynthetic photon flux density (PPFD) of 1200 µmol·m⁻²·s⁻¹, with chamber conditions maintained at an air flow of 600 µmol·s⁻¹, temperature of 25 ± 1 °C, and relative humidity of 60%. Leaf area for P. densiflora was calculated using WinSEEDLE software (ver. 2020a), and indices such as instantaneous transpiration efficiency (ITE) and intrinsic water use efficiency (WUEi) were derived [48,49].

2.5. Statistical Analysis

All statistical analyses were performed using SPSS Statistics 19.0 (SPSS Inc., Chicago, IL, USA). Due to the small sample size of some groups, normality was not assumed, and thus, non-parametric tests were used. Depending on the number of comparison groups, either the Mann–Whitney U test or the Kruskal–Wallis test was conducted with a significance level of p < 0.05. Additionally, principal component analysis (PCA) and biplot visualization were performed using R Studio software (version 4.3.2) to comprehensively explore variable relationships.

3. Results

3.1. Observed External Morphological Features

The cumulative mortality rate of P. densiflora and Q. variabilis affected by surface fires was investigated from 2022 to 2024 (Figure 2). Among the 50 surveyed P. densiflora trees, 6 trees (12.0%) died in 2022, and 9 trees (18.0%) perished in both 2023 and 2024, resulting in a cumulative mortality rate of 18.0% by 2024. In contrast, no crown dieback was observed in Q. variabilis affected by surface fires up to 2024.

The tree characteristics of P. densiflora were recorded as follows: a tree height (TH) of 11.9 ± 2.1 m, a bole length (BL) of 6.5 ± 2.7 m, a diameter at breast height (DBH) of 23.2 ± 5.8 cm, and a circumference at measurement point (CircMP) of 885.0 ± 213.6 mm (

Table 3. Tree characteristics of P. densiflora and Q. variabilis.

Species	Status	TH	BL	DBH	CircMP	BSI
		(m)	(m)	(cm)	(mm)
P. densiflora	Total	11.9 ± 2.1 ***	6.5 ± 2.7 **	23.2 ± 5.8 **	885.0 ± 213.6 **	4.8 ± 1.7 *
	Fine	12.2 ± 1.4 ns	6.7 ± 2.7 ns	23.5 ± 5.8 ns	913.9 ± 201.0 ns	4.5 ± 1.7 *
	Withered	10.4 ± 3.3	5.4 ± 2.1	22.0 ± 5.5	753.3 ± 219.8	6.1 ± 1.7
Q. variabilis	Total	9.3 ± 1.4	4.6 ± 1.1	19.3 ± 2.8	721.7 ± 116.1	3.9 ± 2.1

The values are presented as mean ± SD. Asterisks indicate statistically significant differences according to the Mann–Whitney U-test (* p < 0.05, ** p < 0.01, *** p < 0.001), while “^ns” denotes non-significance. Refer to Table 1 for abbreviated terms.

). These values were approximately 20.0%–38.6% greater than those of Q. variabilis within the same stand (p < 0.05). The bark scorch index (BSI), which quantifies the extent of charring, was 4.8 ± 1.7 in P. densiflora, approximately 25.0% higher than in Q. variabilis (p < 0.05) (Table 3).

A comparison of tree characteristics between healthy (fine) and withered (withered) P. densiflora individuals revealed that fine trees exhibited 6.9%–25.6% greater growth parameters, with CircMP showing the most pronounced difference. However, these differences were not statistically significant (p > 0.05). In contrast, the BSI of withered trees was approximately 32.9% higher than that of fine trees (p < 0.05).

When comparing bark thickness between P. densiflora and Q. variabilis (Figure 3A), no significant differences were observed in thick bark layers (p > 0.05). However, in thin bark layers, P. densiflora exhibited approximately 35.1% thinner bark than Q. variabilis (p < 0.05). When comparing fine and withered P. densiflora individuals, fine trees exhibited relatively thicker bark, but the difference was not statistically significant (p > 0.05) (Figure 3B).

An analysis of the mortality rates according to growth characteristics and BSI grades (Figure 4) revealed that the mortality rates decreased with greater TH, BL, DBH, and CircMP values. Notably, when the BSI exceeded 4.0, the crown mortality rates increased sharply to 25.0%–26.7%.

3.2. Assessment of Trunk Internal Damage

Sonic tomography (SoT) visualizes variations in sound velocity within tree stems as a two-dimensional velocity distribution map, where a lower sound velocity indicates defects such as decay or cavities [39]. Electrical resistance tomography (ERT) evaluates moisture distribution within tree trunks using electrical currents based on the principle that a higher moisture content reduces electrical resistance, whereas a lower moisture content increases resistance [50].

Based on the crown condition from 2022 to 2024, P. densiflora was categorized into three groups: F-F-F (trees remaining healthy until 2024), F-W-W (trees that died in 2023), and W-W-W (trees that died in 2022) (

Table 4. SoT and ERT images of fire-damaged and non-fire-damaged Q. variabilis and P. densiflora.

Species	FD/NFD	F/W	SoT	ERT
				2022	2023	2024
Pd	FD	F-F-F
		F-W-W
		W-W-W
	NFD	F-F-F			-	-
Qv	FD	F-F-F			-	-
	NFD	F-F-F			-	-

Refer to Table 1 for the abbreviated terms.

). SoT did not show significant differences among groups, and all trees exhibited relatively intact density distributions (Table 4). In contrast, ERT displayed distinct differences depending on species, tree condition, and year, prompting further analysis of the color area ratio (CAR) and resistivity (Rs).

In 2022, Q. variabilis exhibited higher overall electrical resistance than P. densiflora, though the difference was not statistically significant (p > 0.05) (

Table 5. ERT color area ratio in P. densiflora and Q. variabilis at 2022.

Species	FD/NFD	ERT Color Area Ratio (%)
		ERTR	ERTY	ERTRY	ERTB
P. densiflora	FD	8.9 ± 9.9 ns	17.8 ± 13.2 ns	26.8 ± 21.0 ns	73.2 ± 21.0 ns
	NFD	5.8 ± 3.4	13.7 ± 8.7	19.5 ± 9.0	80.5 ± 9.0
Q. variabilis	FD	9.4 ± 4.0 ns	21.1 ± 5.9 ns	30.5 ± 7.6 ns	69.5 ± 7.6 ns
	NFD	8.8 ± 3.1	25.7 ± 7.2	34.5 ± 5.8	65.5 ± 5.8

The values are represented in the format of mean ± SD. ^ns denotes non-significance according to the Mann–Whitney U-test. Refer to Table 1 for the abbreviated terms.

). Both species showed higher resistance in fire-damaged trees compared to non-fire-damaged trees. Notably, the ERT_R area ratio in P. densiflora fire-damaged trees was approximately 1.5 times that of non-fire-damaged trees, though this difference was also not statistically significant (p > 0.05).

A temporal analysis of ERT color area ratios (Figure 5) revealed that the ERT_R area ratio increased from 8.9 ± 9.8% in 2022 to 17.5 ± 16.6% in 2023 before decreasing to 11.0 ± 6.4% in 2024. The most pronounced increase occurred in 2023, with greater inter-individual variability. The ERT_R area ratio in 2023 was approximately 1.6–2.0 times higher than in other years, while the ERT_B area ratio decreased by approximately 1.3 times. Resistivity followed a similar trend, with Rs_Min and Rs_Max values in 2023 being 12.9%–52.1% higher than in other years (p < 0.05) (Figure 5).

An analysis of mortality rates based on ERT color area ratios (Figure 6) indicated that higher ERT_R and ERT_Y area ratios correlated with increased tree mortality, particularly when the ERT_R area ratio exceeded 15.0%, at which point the mortality rate surged to 37.5%. Conversely, trees with broader ERT_B area ratios exhibited lower mortality rates, although when the ERT_B area ratio fell below 55.0%, mortality increased to 42.9%.

When examining the changes in the ERT color area ratio across groups classified by canopy health status from 2022 to 2024, the F-F-F group consistently exhibited an area ratio in the order of ERT_B > ERT_Y > ERT_R each year, with relatively stable electrical resistance variability compared to the other groups (Figure 7).

In contrast, the F-W-W group displayed higher ERT_R and ERT_Y area ratios overall compared to the F-F-F group in 2022, the year preceding mortality, leading to relatively higher electrical resistance values (p < 0.05). In 2023, when mortality was determined, the ERT_R area ratio was significantly higher than that of the W-W-W group, with the largest inter-individual variation observed. Additionally, the W-W-W group, determined to have died by 2022, exhibited generally high ERT_R, ERT_Y, and ERT_RY area ratios. An individual-level analysis confirmed that electrical resistance remained consistently high throughout the period (Figure 7). This trend was also evident in the resistivity analysis, where the F-W-W group exhibited elevated resistivity from 2022, the year prior to mortality. Notably, in 2023, the Rs_Min value of the F-W-W group was statistically higher than in other years (p < 0.05) (Figure 7). While no significant differences were detected in the stand-level analysis, clear differences emerged at the individual level, highlighting the necessity of individual-based assessments for detecting subtle physiological changes.

3.3. Physiological Characteristics of the Canopy Layer

A comparison of the chlorophyll pigment content in fire-damaged (FD) and non-fire-damaged (NFD) trees (

Table 6. Chlorophyll and carotenoid contents of fire-damaged and non-fire-damaged P. densiflora and Q. variabilis.

Treatment	Chlorophyll (mg·g−1)			Carotenoid	Chl a/b	T Chl/Car
	a	b	Total	(mg·g−1)
PdFD	8.1 ± 1.1 *	2.8 ± 0.6 ns	10.9 ± 1.4 *	2.2 ± 0.4 *	3.0 ± 0.5 ns	5.1 ± 1.0 ns
PdNFD	11.5 ± 1.1	3.5 ± 0.3	15.0 ± 1.4	2.9 ± 0.1	3.3 ± 0.00	5.3 ± 0.3
QvFD	24.6 ± 2.4 ns	12.5 ± 2.3 ns	37.1 ± 4.5 ns	5.9 ± 0.7 ns	2.0 ± 0.2 ns	6.4 ± 1.1 ns
QvNFD	26.8 ± 0.9	14.2 ± 0.6	40.9 ± 0.4	6.5 ± 0.4	1.9 ± 0.1	6.3 ± 0.3

Values are presented as mean ± SD. Asterisk indicates significant differences at p < 0.05, and ^ns denotes non-significance according to Mann–Whitney U-test. Refer to Table 1 for abbreviated terms.

) revealed that non-fire-damaged P. densiflora and Q. variabilis exhibited higher chlorophyll a and b, total chlorophyll, and carotenoid contents. Specifically, P. densiflora NFD trees exhibited a 25.8%–41.5% higher pigment content than FD trees (p < 0.05), while Q. variabilis showed no significant differences. However, some FD individuals exhibited pigment levels comparable to those of NFD trees, suggesting potential recovery in physiological responses.

Photosynthetic and stomatal responses did not significantly differ between FD and NFD trees (Figure 8). No significant differences (p > 0.05) were observed in parameters related to energy fluxes per reaction center of photosystem II (ABS/RC, DIo/TC, and TRo/RC), the PSII donor-side limitation indicator (V_K/V_J), or the quantum yield of electron transport (φPo, ψEo, and φEo). Additionally, indices reflecting the functional vitality of the photosynthetic apparatus under environmental stresses, such as PI_ABS, DF_ABS, and SFI_ABS, also showed no significant differences. Likewise, key indicators of photosynthetic capacity, including the maximum photosynthetic rate, stomatal conductance, and water use efficiency, did not differ significantly between the two groups.

3.4. Cross-Species Comparison of Fire Damage Indicators

A comparative analysis of post-fire responses between P. densiflora and Q. variabilis reveals species-specific differences in survival, structural protection, and physiological resilience (

Table 7. Summary of post-fire physiological and morphological responses in P. densiflora and Q. variabilis.

Variable/Indicator	P. densiflora (Fire-Damaged)	Q. variabilis (Fire-Damaged)	Key Observation/Interpretation
Mortality Rate (2022–2024)	18.0% cumulative mortality	0% mortality	Higher post-fire survival in Q. variabilis
Bark Scorch Index (BSI)	High BSI in dead trees	Significantly lower BSI	BSI closely linked with tree mortality in P. densiflora
Bark Thickness	Generally thinner	Thicker bark	Thicker bark may have protected Q. variabilis from heat damage
Internal Assessment (SoT/ERT)	No clear internal decay (SoT); varied ERT values; higher ERTR ratio in trees that withered the following year	Minimal damage patterns in both SoT and ERT	Early internal damage detected using ERT
Canopy Condition	Variable, sometimes linked with BSI	Stable, healthy canopy	Crown stability higher in Q. variabilis post-fire
Physiological Response	Decreased pigment content, but no significant impact on photosynthetic activity (A, gs, and PIABS)	Stable pigment levels and physiological traits	Pigment degradation occurred in P. densiflora, but function remained intact

Refer to Table 1 for the abbreviated terms.

). P. densiflora experienced an 18.0% cumulative mortality rate from 2022 to 2024, with tree death strongly associated with higher bark scorch index (BSI) values and thinner bark. In contrast, Q. variabilis recorded 0% mortality and demonstrated stable canopy conditions, likely supported by its thicker bark and lower BSI, which may have provided enhanced protection against thermal injury.

Physiologically, P. densiflora exhibited pigment degradation post-fire, but photosynthetic functions—including A, gs, and PI_ABS—remained largely unaffected. Internal assessments using sonic tomography (SoT) and electrical resistance tomography (ERT) further distinguished the species: while SoT failed to detect clear signs of decay in either species, P. densiflora trees that withered the following year showed elevated ERT_R ratios, indicating early functional decline detectable via ERT. Meanwhile, Q. variabilis exhibited minimal internal damage and maintained physiological stability. These findings underscore the role of anatomical and functional traits—such as bark thickness, internal conductivity, and crown integrity—in shaping post-fire resilience, with Q. variabilis demonstrating markedly superior adaptive capacity under fire stress.

3.5. Results of Principal Component Analysis (PCA)

According to the PCA results of fire-damaged P. densiflora trees (Figure 9), PC1 and PC2 accounted for 31.3% and 20.7% of the total variance, respectively, explaining approximately 52.0% of the overall variability. PC1 was strongly associated with electrical resistance-based indices, including ERT_R, ERT_Y, ERT_RY, and ERT_B, while PC2 was closely related not only to resistivity indices such as Rs_Min and Rs_Max but also to morphological parameters, including BSI, H, and DBH. In terms of population distribution, the F-F-F group was positioned near the center of the PCA dimension, whereas the F-W-W and W-W-W groups tended to occupy extreme positions along specific axes.

4. Discussion

4.1. Interpretation of Morphological Changes

Topographical factors, such as steep slopes and stand density, as well as meteorological conditions like strong winds, influence wildfire spread, intensity, and duration, potentially leading to severe fires [51,52,53,54,55]. Since the examined Pinus densiflora and Quercus variabilis were exposed to the same wildfire within the same site, the observed differences in damage are more likely attributable to individual tree characteristics and interspecific differences in fire resistance rather than variations in site conditions or weather.

Kwon et al. (2021) [28] reported that three years after surface fire damage, the cumulative mortality rate of P. densiflora trees ranged from 22.0% to 27.9%, increasing to 24.9%–32.6% the following year. Similarly, in this study, the cumulative mortality rate of bark-scorched P. densiflora trees increased to 18.0% by 2023, approximately three years post-fire. However, no additional tree mortality was observed in 2024, suggesting that the stabilization of fire-damaged trees had occurred (Figure 2).

Tree mortality following fire damage is closely associated with various tree traits, including the total height, diameter at breast height (DBH), wood density, height of bark charring, and canopy condition, as well as species-specific characteristics such as cork layer thickness, sapwood depth, and volatile compound content [56,57,58]. A comparison of tree characteristics between P. densiflora and Q. variabilis, which experienced similar surface fire damage (Table 3), showed that despite P. densiflora exhibiting higher values for tree height (TH), bole length (BL), DBH, and circumference at measurement point (CircMP), the species still exhibited a cumulative mortality rate of 18.0% (Figure 2). This finding indicates that tree survival is not solely determined by morphological attributes but is significantly influenced by species-specific fire resistance traits.

Bark thickness plays a crucial role in protecting xylem and phloem tissues from high temperatures and pests and is a key determinant of fire resistance [59,60,61]. In this study, the thin-layer bark of fire-damaged Q. variabilis trees was found to be thicker than that of P. densiflora (Figure 3), which is a major factor contributing to the classification of Q. variabilis as a fire-resistant species [62]. Conversely, several factors increase the susceptibility of P. densiflora to fire, including its dense canopy structure, thin bark, and the presence of resin-derived terpenoids, which promote both horizontal and vertical fire spread within pine forests [63,64,65,66]. This highlights the inherent vulnerability of P. densiflora to wildfires [67].

In general, a greater tree height reduces canopy exposure time to heat, thereby lowering the risk of fire-induced mortality [68,69,70,71]. Similarly, in this study, taller P. densiflora trees exhibited lower canopy mortality rates among bark-scorched individuals (Figure 4). Moreover, when comparing DBH and CircMP, CircMP showed a gradual decrease in mortality rate (7.1%–11.1%) with an increasing circumference, whereas DBH did not exhibit a strong correlation. This discrepancy is likely due to the CircMP measurement being taken at approximately 0.4 m above ground level, making it more reflective of heat damage effects. However, since DBH is a widely used indicator of overall tree growth status, integrating both CircMP and DBH measurements would provide a more comprehensive assessment of bark-scorched tree health.

Finally, the burn severity index (BSI), a key indicator for assessing fire damage, was found to be an important predictor of tree mortality. In this study, lower BSI values were associated with reduced mortality rates (Figure 4), a trend consistent with previous findings [28]. Notably, when the BSI was ≤4.0, the mortality rate was only 5.9%, suggesting that the BSI serves as a valuable diagnostic metric for evaluating wildfire damage.

Overall, tree mortality following wildfire was influenced not only by stand characteristics such as tree height and DBH but also by species-specific traits, including bark thickness, and external morphological indicators such as the BSI. As demonstrated by the differences in mortality rates between P. densiflora and Q. variabilis, species-specific traits play a crucial role in determining fire resistance. Therefore, applying wildfire damage criteria from one species to another has inherent limitations, highlighting the need for species-specific diagnostic indicators.

4.2. Insights from Trunk Damage Diagnosis

The comparison between sonic tomography (SoT) and electrical resistance tomography (ERT) in response to surface charring revealed differing trends (Table 4). SoT did not show significant variation based on surface charring or tree species, generally indicating a uniform density distribution of wood. This suggests that fire intensity, such as surface fire, in trees that have been growing stably in forest environments does not induce immediate physical degradation, such as decay or cavitation, in the short term. In contrast, ERT exhibited significant changes in fire-sensitive species like P. densiflora, indicating that surface fire primarily induces physiological and functional impairments, such as vessel occlusion and cavitation, rather than direct structural damage to the trunk.

These characteristics of SoT and ERT have been corroborated in previous studies. Deflorio et al. (2008) [33] reported that in an inoculation experiment with white and brown rot fungi on various tree species, including Pseudotsuga menziesii and Quercus robur, SoT failed to detect early-stage decay even 27 months post-inoculation. Conversely, Bieker et al. (2010) [35] demonstrated that ERT effectively tracked early decay progression within three months in Fraxinus excelsior inoculated with Trametes versicolor. This highlights the sensitivity of ERT to functional and physiological changes associated with structural alterations, particularly variations in the trunk moisture content and hydraulic conductivity. Therefore, utilizing these characteristics could enhance the effectiveness of diagnosing early-stage internal trunk damage due to fire.

Fire-induced damage primarily manifests as physiological and functional changes due to high temperatures, followed by longer-term structural degradation caused by bark exfoliation and fungal infections. Given this progression, ERT should be prioritized for diagnosing fire damage, with SoT being used complementarily to assess long-term structural changes.

The high ERT_R area ratio observed in fire-damaged P. densiflora indicates internal water transport disruptions and physiological stress, which are likely to appear in trees at a higher risk of mortality. In contrast, fire-resistant species such as Q. variabilis exhibited relatively stable electrical resistance values in sapwood, reflecting the maintenance of vessel conductivity (Table 5). These differences highlight species-specific fire resistance and post-fire recovery capabilities, suggesting that physiological responses and structural damage from fire vary significantly by species. Consequently, fire management and restoration strategies should adopt a species-specific approach, with an emphasis on utilizing ERT for the early diagnosis of high-mortality-risk trees.

A comparison of ERT color area ratios from 2022 to 2024 (Figure 5) revealed an overall increase in electrical resistance in 2023, along with greater inter-individual variability. Notably, the ERT_R area ratio in 2023 was approximately 1.6–2.0 times higher than in 2022 and 2024, while the ERT_B area ratio decreased by approximately 1.3 times. These findings are likely linked to early spring precipitation levels. According to the Korea Meteorological Administration, the cumulative precipitation in Jeongseon, Gangwon State, during early spring (February–April) in 2023 was 65.9 mm, less than half of that in 2022 (125.9 mm) and 2024 (166.8 mm).

Spring is a critical period for tree growth, requiring sufficient water supply. However, reduced precipitation in 2023 may have exacerbated soil moisture deficits and atmospheric dryness, further impairing water transport functions in fire-damaged trees. This effect is particularly pronounced in trees already physiologically weakened by fire damage, where additional environmental stressors may have accelerated electrical resistance increases. These results indicate that persistent post-fire climatic conditions can have long-term effects on the physiological recovery of damaged trees, with subsequent climatic variations potentially exacerbating fire damage. Thus, a more precise analysis of the interactions between post-fire physiological changes and climatic factors is warranted. In conclusion, the annual variations in ERT color area ratios and resistance values reflect the impact of climatic conditions, particularly cumulative precipitation, on the internal physical and physiological properties of trees. This underscores the potential utility of ERT_R and ERT_B area ratios as critical indicators of tree health and viability.

The analysis of ERT color area ratios and mortality rates in fire-damaged P. densiflora (Figure 6) revealed a positive correlation between increased ERT_R and ERT_Y area ratios and tree mortality. These findings suggest that the physiological and structural changes induced by fire damage are reflected in electrical resistance characteristics.

Key physiological changes in fire-damaged trees include cambial damage and decreased hydraulic conductivity. When bark fails to provide sufficient thermal insulation during extreme heat events, internal tissues such as the vascular cambium, xylem, and phloem become exposed to high temperatures, leading to cell damage and necrosis, which restricts water and nutrient transport [72]. If cambial necrosis surpasses a critical threshold, vertical water and nutrient transport is obstructed, reducing hydraulic conductivity and increasing the trunk’s overall electrical resistance [20]. Additionally, post-fire moisture loss and increased embolism formation in xylem vessels can further obstruct water transport, heightening the risk of canopy dehydration and tree mortality [20,29,73]. As these physiological stresses accumulate, trees with high ERT_R and ERT_Y area ratios exhibit a rapid increase in mortality rates. In this study, when the ERT_R area ratio exceeded 15.0%, the mortality rate increased to over 37.5% (Figure 2 and Figure 4), suggesting that this metric could serve as a crucial predictor of physiological damage in fire-damaged trees.

Conversely, a high ERT_B area ratio was associated with the preservation of healthy xylem vessels. The study confirmed that as the ERT_B ratio decreased, mortality rates increased, with mortality reaching 42.9% when the ERT_B ratio fell below 55.0%. This indicates that the structural integrity of internal tissues plays a crucial role in tree survival post-fire.

These results demonstrate that post-fire physiological changes are reflected in electrical resistance characteristics, with cambial necrosis and hydraulic conductivity decline closely linked to ERT indicators. Consequently, ERT analysis proves to be an effective tool for the early diagnosis of mortality risk in fire-sensitive species like P. densiflora and for developing targeted management strategies. Furthermore, by establishing threshold values for each ERT color area ratio, this method can serve as a quantitative tool for assessing physiological integrity and predicting mortality risk in fire-damaged trees.

Differences in ERT color area ratios and resistance values among F-F-F, F-W-W, and W-W-W groups suggest a strong link between canopy mortality and trunk electrical resistance characteristics. Notably, in the F-W-W group, ERT_R and ERT_Y area ratios and Rs_Min values were already significantly higher in 2022 (the year preceding mortality) compared to the F-F-F group (Figure 7). This supports the hypothesis that increasing electrical resistance precedes mortality in high-risk trees, highlighting its potential as an early warning indicator.

Ultimately, this study compared the efficacy of non-destructive SoT and ERT techniques for diagnosing physiological and structural changes in fire-damaged trees and assessing mortality risk. The findings indicate that ERT effectively detects internal physiological stress and structural damage, with ERT_R and ERT_Y area ratios increasing in parallel with mortality risk. This underscores ERT’s potential as a valuable tool for diagnosing early physiological changes and identifying high-risk trees post-fire.

However, ERT showed a tendency to decline after reaching a certain damage threshold, which may be linked to methodological limitations. In severely damaged trees, internal structural degradation and fungal colonization can alter electrical conductivity, as observed in previous decay inoculation studies. Thus, while ERT is useful for detecting early-stage damage, its interpretation in long-term damage progression may be limited. In contrast, SoT, although less effective for early diagnosis, may provide valuable insights into long-term structural changes.

In this study, when analyzing the entire stand, ERT-related indicators showed a slight increasing trend (Figure 5). However, since dead trees accounted for only 18.0% of the total (Figure 2), overall stand-level damage was limited. These results suggest that internal diagnostic assessments using ERT may reflect patterns of individual tree decline, which broadly align with observed mortality. While ERT can be a useful tool for evaluating tree health, decisions regarding post-wildfire restoration should primarily be based on observed mortality rates and site-specific conditions such as soil stability and seed source availability, with ERT serving as supplementary diagnostic data.

To better assess the mortality risk of fire-damaged trees, early screening using ERT can provide valuable insights, particularly due to its sensitivity to physiological changes in the early stages following fire exposure. Complementing this, SoT is useful for evaluating longer-term structural and physical damage. While these diagnostic methods enhance our understanding of internal tree conditions, they should be integrated with direct observations of mortality and site-specific factors. Such a combined approach can contribute to more informed evaluations of forest conditions and assist in shaping appropriate post-wildfire restoration strategies, including the prioritization of recovery efforts.

4.3. The Physiological Responses of the Canopy Layer

O’Brien et al. reported that the chlorophyll content in Pinus palustris needles decreases as canopy scorching increases, primarily due to root and mycorrhizal loss and canopy scorching [74]. The loss of fine roots, which absorb water and inorganic nutrients from the soil, along with their associated mycorrhizae, reduces overall nutrient uptake. Additionally, canopy scorching affects stomatal function and photosynthesis, limiting physiological processes and impeding nutrient transport through the xylem [75]. Consequently, both the efficiency of nutrient translocation from the roots to the canopy and the total amount of nutrients available are reduced, leading to a decreased chlorophyll content.

In this study, pine trees affected by surface fire exhibited significantly lower chlorophyll a, chlorophyll b, total chlorophyll, and carotenoid contents compared to unaffected trees (p < 0.05) (Table 6). This finding suggests that reduced root vitality, soil nutrient loss, and diminished hydraulic conductivity in the stem contributed to the decline in pigment content. In particular, P. densiflora, which has low fire resistance, appears to have experienced more severe physiological stress. Conversely, there were no statistically significant differences in the chlorophyll content between affected and non-fire-damaged Q. variabilis trees, likely due to the species’ higher fire resistance and limited internal stem damage caused by the fire.

Thermal stress generally leads to the inactivation of PSII photochemical reactions or a reduction in electron transport from QA to QB, which also affects the structural integrity of the chlorophyll–protein complex [76,77]. These physiological changes are associated with an increase in minimal fluorescence (F0) and a decrease in maximal fluorescence (FM), leading to a reduction in the maximum quantum yield of primary photochemistry (φPo) [78]. Additionally, high temperatures induce the appearance of the K step in the Kautsky curve, which is linked to the inactivation of the oxygen-evolving complex (OEC) [79]. Increases in ABS/RC and TRo/RC indicate a lower proportion of active reaction centers, meaning that energy capture per active reaction center is elevated. This suggests an accumulation of excessive energy within the photosynthetic apparatus, which can lead to oxidative damage over time [80]. To mitigate oxidative stress, plants increase the number of active reaction centers and dissipate excess energy as heat or other forms of non-photochemical quenching [81,82].

However, in this study, there were no significant differences in the chlorophyll a fluorescence parameter between fire-damaged and non-fire-damaged trees of both species. While certain reaction center-related parameters (V_K/V_J, ABS/RC, TRo/RC, and DIo/RC) showed slight increases in fire-damaged trees, these differences were not statistically significant (p > 0.05). Furthermore, stress-sensitive indices such as PI_ABS, DF_ABS, and SFI_ABS also exhibited no significant variations (Figure 8).

Additionally, previous studies have reported that the net photosynthetic rates temporarily increased by approximately 30%–50% in fire-damaged trees compared to non-fire-damaged trees, except in Acer rubrum, immediately after fire. Simultaneously, soil nutrient levels, including nitrogen, phosphorus, and potassium, also increased [83,84,85]. Other studies have noted that soil magnesium, manganese, and zinc levels also rise post-fire, and that these changes in soil nutrient availability are closely correlated with physiological processes in plant leaves [86,87]. Renninger et al. investigated the physiological responses of Pinus rigida before and after prescribed fire, revealing that sap flow rates in fire-damaged trees decreased by approximately 27% compared to non-fire-damaged trees, a trend that persisted for five to six months [75]. While the maximum photosynthetic rates initially increased after the fire, they declined during August, when temperatures reached approximately 31.7 °C, due to sustained reductions in stomatal conductance. Although the initial increase in photosynthetic rates may have resulted from a temporary rise in soil fertility [88,89,90], the prolonged decrease in sap flow and the combined effects of high summer temperatures likely contributed to the subsequent reduction in maximum photosynthesis.

In this study, comparisons of key photosynthetic and stomatal parameters including the net photosynthetic rate, stomatal conductance, transpiration efficiency, and water use efficiency between visibly unaffected FD and NFD trees revealed no statistically significant differences (p > 0.05) (Figure 8). The absence of significant differences in both chlorophyll a fluorescence responses and gas exchange parameters suggest that, over time, understory vegetation and forest ecosystems have stabilized following the fire, and that post-fire rainfall may have leached inorganic nutrients from the soil [85,91]. These findings indicate that stem-level assessments, rather than canopy analyses, may be more effective in evaluating internal fire-induced damage in trees.

Conclusively, FD trees with visually intact canopies exhibited stable physiological functions, as shown by the photosynthetic and stomatal responses, chlorophyll content, and chlorophyll a fluorescence parameters. These results indicate that canopy-based assessments alone may not adequately detect underlying damage or predict long-term mortality, as FD trees in this study showed canopy health comparable to that of NFD trees. However, stem-level damage may compromise the tree’s ability to cope with future environmental fluctuations such as drought, temperature extremes, or intense rainfall. Although the long-term survival of many pine species with basal fire scars has been documented, it remains unclear whether P. densiflora exhibits a similar pattern. Therefore, long-term monitoring using tools like ERT could provide important insights into delayed mortality risks and support more accurate post-fire management decisions.

4.4. Implication from Principal Component Analysis (PCA)

The principal component analysis (PCA) conducted in this study revealed that while some fire-damaged P. densiflora exhibited a tendency to be distinguished based on their physiological and morphological characteristics, clear boundaries between groups were not observed overall (Figure 9). However, considering the key variables constituting the principal component axes, PC1 was primarily explained by color-area-based indices such as ERT_R and ERT_Y area ratios. This suggests that PC1 serves as a major axis reflecting physiological stress and functional anomalies within trees caused by fire damage. In other words, fire-damaged trees experience changes in stem moisture content due to tissue damage and functional loss in the stem, which in turn alters electrical resistance values. This indicates that such variations are among the primary factors influencing inter-individual differences.

Meanwhile, PC2 was influenced not only by resistance-related indices such as Rs_Min and Rs_Max but also by external morphological variables such as the BSI, H, and DBH. This suggests that PC2 represents a more comprehensive axis encompassing not only physiological responses but also tree stand characteristics and external morphological features (Figure 9). Specifically, while PC1 emphasizes ERT-based color area ratios reflecting damage patterns in the stem, PC2 integrates physiological stress with structural and morphological traits of the tree.

When examining the distribution of individuals, the F-F-F group was mostly located at the center of the PCA space (Figure 9), representing an average inclusion of various variables. In contrast, the F-W-W and W-W-W groups exhibited extreme positions along certain axes, implying that these groups were more severely affected by physiological stress due to fire damage, potentially experiencing greater impairment in stem water transport and tissue integrity.

These findings support the potential utility of ERT-based indices (ERT_R, ERT_Y, ERT_RY, ERT_B, Rs_Min, Rs_Max, etc.) and external morphological characteristics (BSI, H, BL, DBH, CircMP, etc.) in assessing physiological stress and functional anomalies in fire-damaged trees. In particular, analyzing groups exhibiting extreme values along specific PCA axes may facilitate the early identification of high-risk trees, contributing to the development of effective management strategies.

5. Conclusions

In general, trees with larger stature—such as greater DBH and height—are known to exhibit higher fire resistance, and this study confirmed a similar trend: trees with better stand traits showed lower mortality rates. Additionally, this study demonstrated the utility of the bark scorch index (BSI) as a key indicator for predicting the mortality of Pinus densiflora following surface fire damage. However, fire damage is not solely determined by stand characteristics and the BSI; species-specific traits such as bark thickness and flammability also play a crucial role in determining the severity of damage.

This pattern was especially evident in P. densiflora, which is highly susceptible to fire. Despite exhibiting larger DBH and height—20.0% and 38.6% greater, respectively—compared to Quercus variabilis, P. densiflora had a cumulative mortality rate of 18.0%. This higher vulnerability was likely due to its thin bark and high stem electrical resistance. In contrast, Q. variabilis, though it possessed relatively smaller DBH and height, exhibited no mortality. Its thick bark and cork tissue effectively insulated the sapwood, preserving xylem function and resulting in low stem electrical resistance and stable canopy physiological responses.

Wildfire-induced tree mortality typically begins with internal stem defects caused by heat, which impair water transport and may lead to death. This study compared sonic tomography (SoT) and electrical resistance tomography (ERT) to diagnose such changes. ERT was shown to be highly sensitive to both physiological stress and structural damage, correlating closely with observed canopy mortality. Notably, in the F-W-W group, increases in ERT_R and ERT_Y area ratios and Rs_Min values in 2022 (the year preceding mortality) clearly distinguished high-risk trees from those in the F-F-F group. These results suggest that elevated electrical resistance may serve as an early warning signal for impending mortality.

While SoT did not detect significant changes in the early stages, it remains a valuable tool for assessing long-term physical changes in woody tissue. A combined approach is therefore recommended: prioritize ERT for early physiological screening and use SoT to monitor long-term structural changes in fire-damaged stems.

Stem damage, which disrupts the connection between the roots and canopy, can impair hydraulic conductivity and reduce the internal water content, ultimately destabilizing the tree’s physiological functions. However, in this study, trees maintained stable canopy conditions despite fire-induced stem damage, indicating that their photosynthetic functions were not immediately compromised. This finding emphasizes the importance of assessing stem integrity to more accurately diagnose and manage fire-damaged trees.

In conclusion, this study presents a comprehensive methodology for diagnosing the physiological and structural conditions of surface-fire-damaged trees using a combined analysis of the BSI, stand characteristics, and internal stem diagnostics. These findings offer a scientific foundation for the early identification of high-risk individuals and more effective post-fire restoration planning, thereby contributing to the resilience and long-term health of fire-affected forest ecosystems.