Bark Browsing and Recovery: A Comparative Study between Douglas Fir and Silver Fir Species in the Western Carpathians

Abstract

The foraging behavior of ruminating ungulates significantly impacts forest ecosystems due to their nutritional requirements. This study focuses on inter-specific comparisons of bark browsing (stripping) between Douglas fir (Pseudotsuga menziesii (Mirb.) Franco) and silver fir (Abies alba Mill.). Field measurements were conducted at the previously established research demonstration site “Husárik” in the Javorníky Mountains, north-western Slovakia. We measured the sizes of wounds on stems and assessed the dimensions of unrecovered parts of wounds, subsequently calculating areas of recovered bark patches. Then, the total areas of wounds recovered patches, and unrecovered parts were measured. Additionally, the percentages of wound areas, recovered areas, and unrecovered areas were expressed as proportions relative to the total bark surface. Influencing factors such as the position of wounds along the vertical stem profile and stem diameter class were analyzed using two- and three-way ANOVA followed by Fisher’s LSD test. The results demonstrate the susceptibility of both tree species to bark browsing by ruminating ungulates, primarily red deer (Cervus elaphus L.) in our study site. Douglas fir exhibited slightly less intense browsing compared to silver fir, while silver fir had a slightly higher recovery rate. Specifically, 17.0% of the bark surface in Douglas fir and 21.5% in silver fir were browsed. The proportions of recovered areas on wounds were 62.5% and 69.6% in Douglas fir and silver fir, respectively. Regarding the vertical stem profile, the most intense browsing occurred at a height of 101–125 cm from the ground level, with rates of 40.7% in Douglas fir and 47.0% in silver fir. Browsing intensity showed negligible variation among diameter classes, which were classified as up to 35 mm, 35–50 mm, and over 50 mm. Our findings suggest that, as an introduced species, Douglas fir is not a suitable substitute for other commercially significant tree species in supporting sustainable forestry in European countries where ruminating ungulates are overabundant.

1. Introduction

Forests play a crucial role in providing a diverse array of ecosystem services essential for human well-being. The sustainability of forests is paramount in fulfilling this mission, hinging upon the maintenance of continuous forest cover [1] and the prevention of tree damage and stand disturbances [2]. The vitality of forest stands, encompassing aspects such as ecosystem services, biodiversity, and long-term wood yield within the framework of sustainable development at the stand level, relies significantly on effective management during the stages of forest regeneration and young growth. Therefore, these critical stages must be safeguarded against both abiotic and biotic threats (e.g., [3]). Outputs from the Slovak National Forest Inventory highlight ruminating ungulate game as the most prevalent harmful agent in forest stands of the first age class [4], which is related to the overabundance of this game group. Actually, a high population density of ruminating ungulate game is not unique only to Slovakia but is observed in numerous European countries [5,6,7,8]. Consequently, the overabundance of game poses a significant challenge to the sustainable management of forests, both at present and in the foreseeable future across Europe.

The findings from the Slovak National Forest Inventory [4], for instance, revealed that the trees most impacted by game browsing extended beyond softwood broadleaved species, which typically have lower economic contributions, even commercial species like the silver fir (Abies alba Mill.). The silver fir, in particular, often experiences intense damage from ruminating ungulates, especially during its initial growth stages. Seedlings are commonly damaged by the biting off of shoots, while saplings or larger trees face harm mostly through stem bark browsing [9]. Several authors [10,11] have demonstrated that game browsing can lead to a significant decrease in species composition and, in many cases, the complete disappearance of the silver fir. This is a critical concern given the importance of the silver fir as a tree species in many Eurasian regions from both production and ecological perspectives [12]. For example, the silver fir plays a vital role in mixed forest stands by improving soil conditions [13], enhancing species biodiversity [14], and supporting wood productivity [15]. Given these ecological and economic benefits, actively managing silver fir through forest management practices [16] and implementing game management strategies, even if it involves reducing game population density [17], becomes imperative.

While silver fir is an autochthonous species in Europe [12], Douglas fir (Pseudotsuga menziesii (Mirb.) Franco) originates from North America and was introduced to Europe [18]. This species has a rather similar habitus (branching, needle shape, bark surface structure, etc.) but significantly different ecological and production properties compared to silver fir. According to Podrázský et al. [19], Douglas fir is the most commonly utilized introduced tree species in the temperate zone, planted in many European countries. Since both silver fir and Douglas fir are more resistant to drought stress than Norway spruce (Picea abies Karst. L), they might be suitable substitutes for spruce-dominant forests under ongoing climate change [20]. However, for Slovakia, a country with an overabundance of ruminating ungulates, a crucial task might be determining whether Douglas fir is less susceptible to browsing by game compared to other tree species that could serve as substitutes.

In the countries of Central and Western Europe, red deer (Cervus elaphus L.) are the biggest ruminating ungulates and generate the most harm to forests through browsing [6]. In these regions, the second and third most frequent ruminating ungulates are roe deer (Capreolus capreolus L.) and fallow deer (Dama dama L.). However, since the weight of fallow deer and roe deer are about three times and six times lower, respectively, than that of red deer, they have much lower demands for forage quantity. Therefore, they cause much less damage to forest stands [21]. Hence, in Slovakia and other Central European countries, red deer are the primary contributors to bark browsing (stripping) in forest trees. Bark stripping can result in significant economic losses due to wood deformations and/or infections from wood-decaying fungi [22,23]. Beyond these direct consequences, tree stems affected by bark stripping and subsequent rotting are susceptible to snow or wind breakages.

Previous studies have predominantly focused on the evaluation of stem bark browsing, employing simple quantification methods such as the measurement of the area of browsed stem bark at a tree level [24] or the rate of browsed trees at a stand level [25]. Another attempt was made to assess the effects of bark stripping on timber production in Norway spruce [26]. However, we could not find any work related to bark recovery after stripping, much less any findings on inter-specific differences in this recovery process (e.g., time duration of complete recovery and impact on timber quality after recovering) for Douglas fir or silver fir.

The primary objective of our study was to conduct inter-specific comparisons of bark damage intensity caused by game browsing between Douglas fir and silver fir. Additionally, we aimed to extend these comparisons to assess bark recovery after stripping in both tree species. Furthermore, our focus included the analysis of certain tree characteristics, such as stem thickness and position within the stem vertical profile, to understand bark browsing intensity in both Douglas fir and silver fir.

2. Materials and Methods

2.1. Study Site

Our field study took place at the research–demonstration site known as “Husárik” (hereinafter referred to as the Husárik site), situated within the Javorníky Mountains in the Kysuce region of north-western Slovakia. The climatic conditions at the Husárik site are characterized by cold and humid weather, with an average temperature of approximately 16 °C in July and nearly −5 °C in January [27]. The mean annual temperature registers at 6 °C, annual precipitation reaches nearly 1100 mm, and snow cover lasts for around 90 days per year. The bedrock in the Javorníky Mountains is formed predominantly from clay stones and shales with modally acidic, mostly clayey soil. The Husárik site is classified as an Abieto–Fagetum forest type [28]. This classification indicates that the natural forests in this area primarily comprise European beech (Fagus sylvatica L.) with silver fir.

At the Husárik site, akin to numerous locations within the Javorníky Mountains, secondary spruce-dominated monocultures have experienced a notable decline, particularly since the onset of the current century. This decline can be predominantly attributed to climatic extremes, exemplified by the exceptionally dry conditions observed in 2003 and the subsequent outbreaks of bark beetles. Consequently, in 2010, the mature stands at the Husárik site were subjected to “incidental felling,” which involved the removal of trees that had already died, become weakened, or were affected by insect and fungal infestations.

The Husárik site belongs to nearly 3000 hectares of a hunting ground; the area comprises 2/3 of forest and 1/3 of crop fields and grasslands. According to records from the local hunting association, the very early spring (pre-parturition) of 2010 revealed the presence of fifty red deer, sixty roe deer, fifty mouflons, and fifty fallow deer within the hunting ground. In the spring of 2023, the counts increased to thirty-five red deer, fifty roe deer, thirty mouflons, and sixty fallow deer. A study by Červený et al. [9] indicated that roe deer consume approximately a quarter, mouflons a third, and fallow deer half of the forage eaten by red deer. Considering these inter-specific comparisons within the studied hunting ground, it can be estimated that red deer account for 40%, roe deer for 14%, mouflons for 11%, and fallow deer for 35% of the total forage consumed by ruminating ungulates. Although this is a very simplified estimation due to the varied diet composition of each species, it highlights that three-quarters of the consumed forage is attributed to red deer and fallow deer combined. Moreover, the proportion of bark browsing is likely higher for these two species, as roe deer exclusively browse on foliage and shoots. In spite of the fact that mouflons also occasionally browse on stem bark, their forage demands in the studied area are considerably lower than those of red and roe deer, making their contribution to total stem bark browsing negligible (see also [29]).

2.2. Stand Description

At the beginning of 2011, a series of experiments were initiated to evaluate various reforestation and silvicultural strategies. The experiments included exclusive natural regeneration, tree planting, and combinations of both, with varying proportions of natural regeneration and planting. Additionally, different combinations of tree species were explored [30]. In 2011, a protective fence was constructed around a five-hectare section of the Husárik research–demonstration site dedicated to artificial regeneration experiments. This fence, standing at a height of 2.4 m with a mesh size of 45 × 45 mm, effectively restricted, for certain periods, access by ruminating ungulates to the enclosed area. For our measurements, we focused on a portion of the enclosure, representing the plantation of two coniferous species, specifically Douglas fir and silver fir, with an equal distribution of both species (1:1 ratio). The targeted forest complex occupied approximately 0.2 hectares.

The scenario involving the exclusion and subsequent presence of ungulates within the fenced area necessitates a detailed explanation. Approximately seven years after the fence’s construction, some sections of it were damaged. As a result, sporadic sightings of red deer, fallow deer, roe deer, and mouflons occurred within the enclosed space. Despite diligent repairs to the fence in 2019 and 2021, additional damage was sustained between 2019 and 2022. Our assessment suggests that the upper portion of the fence was primarily damaged by red deer, while the lower section bore evidence of wild boar activity (Sus scrofa L.). It appeared that the wild ungulates discerned differences in tree species composition between the stands outside the fence, where less appealing species for nutrition, such as Douglas fir and silver fir, were prevalent. The game, particularly red deer and wild boar, likely “learned” methods to breach the fence and gain access. Consequently, this altered situation provided an opportunity to estimate stem bark browsing on young trees (12 years old) over a brief period spanning three years under heightened population pressure from ungulates. In practice, our initial observations revealed instances of bark browsing during the winter seasons of 2019/2020, 2020/2021, and 2021/2022, but not any in 2022/2023 (no fresh wounds were recorded within the observations performed in autumn 2023).

2.3. Measurements and Analyses

In autumn 2023, we randomly selected 100 Douglas firs and 110 silver firs from the entire stand area, which were then subjected to our measurements. Specifically, tree heights were measured using a hypsometer TruPulse 360° R (Laser Technology, Inc., Centennial, CO, USA) with a precision of ±0.1 m. Simultaneously, stem diameter d_1.3 (situated 130 cm from the ground level) was measured with a digital caliper Masser BT (Masser Precision, Rovaniemi, Finland) with a precision of ±0.1 mm.

Subsequently, we divided the vertical profile of each stem into 25 cm-long sections marked by chalk. The diameters at the lower and upper borders of each section were measured with digital calipers at a precision of ±0.1 mm. Each stripped (browsed) area was measured, specifically its vertical dimension with digital calipers (precision of ±0.1 mm) and its horizontal dimension with a rolling measuring tape (±1.0 mm). If the browsed area extended across stem sections, it was proportionally divided between them. Similarly, we measured the area of the unrecovered part of the wound, representing the inner part of the browsed area (see Figure A1 in Appendix A). Subsequently, the recovered (healed) area was calculated for each wound as the difference between the wound area and the area of the unrecovered part.

Then, the total surface of stem bark was calculated for each stem section using the formula for the surface of the truncated cone excluding the areas of bottom and top cross-sections, i.e.,

$$S_b=\pi \left(r_1+r_2\right)\sqrt{{(r_1-r_2)}^2+l_s^2}$$

where:

S_b is the stem bark surface (cm²),

r₁ is the radius of the bottom end (cm),

r₂ is the radius of the top end (cm),

l_s is the length of the section (i.e., 25 cm).

Later, the total surface of the stem bark along the entire targeted profile, up to a height of 250 cm, was calculated as the sum of surfaces in 25 cm-long consecutive sections (i.e., 0–25 cm, 26–50 cm, 51–75 cm, 76–100 cm, 101–125, 126–150 cm, 151–175 cm, 176–200 cm, 201–225 cm, and 226–250 cm from the ground level).

Afterward, seven stem bark browsing metrics were used. These, expressed in square centimeters, included the browsed areas of bark (abbreviated as BAB), recovered areas of bark (RAB), and unrecovered areas of bark (UAB, see also

Table 1. The indicators (with abbreviations and units) were measured on the wounds in Douglas fir and silver fir trees in the Husárik site, Kysuce region, Slovakia.

Level	Stem Sections		Entire Stem
Indicator	Abbreviation	Unit	Abbreviation	Unit
Browsed area of bark	BABss	cm2	BABes	cm2
Recovered area of bark	RABss	cm2	RABes	cm2
Unrecovered area of bark	UABss	cm2	UABes	cm2
Browsed area to total area	BATAss	%	BATAes	%
Recovered area to total area	RATAss	%	RATAes	%
Unrecovered area to total area	UATAss	%	UATAes	%
Recovered area to browsed area	RABAss	%	RABAes	%

). The other four indicators were expressed as percentage matrices: browsed area of bark to total bark area (BATA), recovered area of bark to total bark area (RATA), uncovered area to total bark area (UATA), and recovered bark area to browsed area (RABA, summarized in Table 1).

Practically, we calculated all the indicators at the stem section level (marked by the lower index “ss”) and at the entire stem profile (lower index “es”, also shown in Table 1), considering a maximum distance from the ground of 250 cm. Data analyses were performed in Excel and Access programs. Since the Kolmogorov–Smirnov test confirmed the normality of the data, we might consider conducting a two-way ANOVA and a three-way ANOVA, followed by a Fisher’s Least Significant Difference (LSD) test. The tests were performed in STATISTICA 10.0 software. The tested factors included tree species (abbreviated as TS), stem diameter classes (DC), and stem section (SS, see Table 1). Regarding stem diameter classes, they were intentionally categorized to ensure at least 30 individuals in each category, specifically classified as under 35 mm, 35–50 mm, and over 50 mm.

3. Results

The set of Douglas fir represented trees with heights ranging from 2.5 to 7.8 m (with a mean of 4.0 m) and diameters at breast height (d_1.3) between 16.5 and 85.0 mm (with a mean of 45.1 mm). At the same time, silver fir trees had heights between 2.4 and 6.0 m (with a mean of 3.6 m) and diameters (d_1.3) ranging from 19.5 to 77.5 mm (with a mean of 42.5 mm). This indicates that both species exhibited very similar tree characteristics (

Table 2. Descriptive statistics for the basic characteristics in the measured Douglas fir and silver fir trees in the Husárik site, Kysuce region, Slovakia.

Tree Species	Characteristics	Mean	S.D.	Minimum	Maximum	25th Percentile	75th Percentile
Douglas fir	Tree height (m)	4.01	1.17	2.50	7.80	3.10	4.60
(n = 100)	Diameter d1.3 (mm)	45.13	18.08	16.50	85.00	30.50	59.00
Silver fir	Tree height (m)	3.63	0.80	2.40	6.50	3.00	4.30
(n = 110)	Diameter d1.3 (mm)	42.53	13.11	19.50	77.50	32.50	52.00

).

Our results showed that the mean values of BAB_es were 556 cm² (standard error ±60 cm²) and 652 cm² (±46 cm²) in Douglas fir and silver fir, respectively (see Figure 1). Although the difference was not large, it was statistically significant (

Table 3. Results of two- and three-way ANOVA for indicators (related to an area in cm², see also Table 1) measured on the wounds for the entire stem level (marked by lower index “es”) and the stem section (marked by lower index “ss”) level in Douglas fir and silver fir trees. TS means tree species, DC is diameter class, and SS is stem section.

Type of ANOVA	Factor vs. Indicators	BABes			RABes			UABes
		Df	F-Value	p-Values	Df	F-Value	p-Values	Df	F-Value	p-Values
two-way	TS	1	13.611	<0.0001	1	34.310	<0.0001	1	2.503	0.115
	DC	2	87.327	<0.0001	2	95.967	<0.0001	2	4.156	0.017
	TSxDC	2	1.442	0.239	2	3.419	0.035	2	3.937	0.020
Type of ANOVA	Factor vs. indicators	BABss			RABss			UABss
		Df	F-value	p-values	Df	F-value	p-values	Df	F-value	p-values
three-way	TS	1	29.436	<0.0001	1	88.197	<0.0001	1	5.138	0.024
	DC	2	188.858	<0.0001	2	241.102	<0.0001	2	8.532	<0.0001
	SS	9	405.010	<0.0001	9	300.746	<0.0001	9	97.346	<0.0001
	TSxDC	2	3.118	0.044	2	8.590	<0.0001	2	8.155	<0.0001
	TSxSS	9	8.025	<0.0001	9	12.288	<0.0001	9	4.995	<0.0001
	DCxSS	18	19.675	<0.0001	18	23.680	<0.0001	18	2.423	<0.0001
	TSxDCxSS	18	1.841	0.017	18	2.116	0.004	18	3.164	<0.0001

Explanatory note: Numbers in bold indicate all p-values < 0.05.

). Similarly, a significantly larger value in silver fir than in Douglas fir was found for RAB_es (466 ± 43 cm² vs. 338 ± 43 cm²). Regarding UAB_es, the difference between Douglas fir (218 ± 37 cm²) and silver fir (186 ± 19 cm²) was insignificant.

A statistically significantly larger value was recorded in silver fir (21.5 ± 1.0%) than in Douglas fir (17.0 ± 1.5%) in BATA_es, as well as in RATA_es (14.9 ± 1.1% vs. 10.0 ± 0.9%; see Figure 2;

Table 4. Results of two- and three-way ANOVA for indicators (related to shares in %, see also Table 1) measured on the wounds for the entire stem level (marked by lower index “es”) and the specific stem section level (marked by lower index “ss”) in Douglas fir and silver fir trees. TS means tree species, DC is diameter class, and SS is stem section.

Type of ANOVA	Factor vs. Indicators	BATAes			RATAes			UATAes			RABAes
		Df	F-Value	p-Values	Df	F-Value	p-Values	Df	F-Value	p-Values	Df	F-Value	p-Values
two-way	TS	1	26.469	<0.0001	1	54.712	<0.0001	1	0.382	0.537	1	8.044	0.005
	DC	2	2.402	0.93	2	12.663	<0.0001	2	5.59	0.003	2	7.196	<0.0001
	TSxDC	2	1.102	0.334	2	3.362	0.037	2	1.006	0.367	2	2.877	0.059
Type of ANOVA	Factor vs. indicators	BATAss			RATAss			UATAss			RABAss
		Df	F-value	p-values	Df	F-value	p-values	Df	F-value	p-values	Df	F-value	p-values
three-way	TS	1	56.593	<0.0001	1	126.623	<0.0001	1	0.688	0.407	1	7.587	0.006
	DC	2	12.052	<0.0001	2	46.891	<0.0001	2	10.326	<0.0001	2	37.289	<0.0001
	SS	9	455.791	<0.0001	9	344.52	<0.0001	9	102.707	<0.0001	9	227.703	<0.0001
	TSxDC	2	3.196	0.041	2	10.387	<0.0001	2	3.002	0.049	2	1.855	0.157
	TSxSS	9	12.120	<0.0001	9	16.94	<0.0001	9	4.330	<0.0001	9	4.805	<0.0001
	DCxSS	18	3.931	<0.0001	18	4.832	<0.0001	18	3.625	<0.0001	18	1.588	0.055
	TSxDCxSS	18	0.868	0.0619	18	2.137	0.004	18	1.395	0.123	18	1.089	0.356

Explanatory note: Numbers in bold indicate all p-values < 0.05.

). At the same time, very small inter-specific differences were found for UATA_es, with the value being 7.0% (±1.3%) in Douglas fir and 6.6% (±0.7%) in silver fir. Additionally, a significantly larger value of RABA_es was observed in silver fir (69.6 ± 2.5%) than in Douglas fir (62.5 ± 4.6%).

Since BAB and BATA are the most illustrative indicators of bark browsing, we analyzed them in more detail, specifically regarding diameter classes and/or stem sections. The results showed that BAB_es increased significantly with the rising diameter class (Figure 3a; Table 3). For Douglas fir, the smallest value of BAB_es was 300 (±59) cm² in the diameter class up to 35 mm, and the largest one was 767 (±91) cm² in the diameter class over 50 mm. Similarly, in silver fir, the minimum value of BAB_es was 300 (±59) cm² in the diameter class up to 35 mm, and the maximum value of 902 (±68) cm² was found in the diameter class over 50 mm. It is noteworthy that the increasing values of BAB_es in both species are logically related to the increasing bark surface with larger tree size (stem diameter). Moreover, considering BAB_es within each individual diameter class, significant inter-specific differences were found for trees in the diameter class up to 35 mm and those in the diameter class over 50 mm (Figure 3a).

While BAB_es showed an increasing tendency with diameter classes, rather equal values among the diameter classes were found for BATA_es (Figure 3b; Table 4). On the other hand, larger values in silver fir than in Douglas fir were recorded in all diameter classes; the only insignificance was observed in the diameter class of 35–50 mm. The minimum value of BATA_es (16.0 ± 3.0%) was found in Douglas fir in the diameter class up to 35 mm, and the maximum value (22.1 ± 1.5%) was in silver fir in the diameter class over 50 mm.

Furthermore, we calculated and analyzed BAB and BATA on a stem section level (see Figure 4a,b; Table 3 and Table 4). We found that the maximum values of both indicators were in stem section no. 5, meaning a distance from the ground of 101–125 cm. The BAB_ss values in this section were 163.9 (±6.5) cm² for silver fir and 150.5 (±8.2) cm² for Douglas fir. Similarly, the BATA_ss reached the maximum values (47.0 ± 1.3% in silver fir and 40.7 ± 1.5% in Douglas fir) in the stem section located 101–125 cm from the ground level. Significant inter-specific differences were found in the stem sections 4–7, i.e., 76–175 cm from the ground.

In addition, we analyzed the combined effect of stem thickness (diameter class) and distance from the ground (stem section) on BAB_ss and BATA_ss (Figure 5; Table 3 and Table 4). The statistical tests clearly showed that BAB_ss differed between the species and was related to the diameter class as well as the stem section. Moreover, all combinations of these factors (tree species, diameter class, and stem section) influenced the values of BAB_ss (Table 3). The maximum values of BAB_ss were in the diameter class over 50 mm in stem section 5, specifically 221.5 (±13.0) cm² in Douglas fir and 224.3 (±13.1) cm² in silver fir. In the case of BATA_ss, the values differed between the species and were also related to the diameter class as well as the stem section. Moreover, combinations of these factors (i.e., tree species × diameter class × stem section) influenced the values of BATA_ss, except for the combination of all three factors (i.e., tree species × diameter class × stem section). Here, the maximum values were in the diameter class over 50 mm in stem section 5, specifically 41.8 (±2.3) % in Douglas fir and 48.6 (±2.1) % in silver fir.

Finally, knowing that the stems of both fir species are very intensively browsed on bark stems within the range of 76–175 cm from the ground level, we attempted to create models of bark surface related to diameter d_1.3 (see Figure 6). The relationships were linear and very close (r² = 0.97 and 0.92 in silver fir and Douglas fir, respectively). Moreover, the fitted lines were very close to each other, indicating that the surfaces of the stems within sections positioned 76–175 cm from the ground level were very similar in both species. The models show that the stem surface for a tree with a diameter d_1.3 of 40 mm was about 1570 cm², while for a tree with a diameter of 80 mm, it was nearly double (about 3130 cm²). These areas represent a certain potential that can be consumed by ruminating ungulates, especially red deer, if the entire surface of the stem in the selected section was browsed.

4. Discussion

4.1. Bark Browsing (Stripping)

Our results showed that BATA_es were 17.0% and 21.5% in Douglas fir and silver fir, respectively. This implies that approximately 1/5 of all bark within the available space (considering red deer), up to 250 cm from the ground, was browsed. This bark browsing occurred over three winter seasons: 2019/2020, 2020/2021, and 2021/2022. Previously, we conducted a similar study at the same site for common aspen (Populus tremula L.), common rowan (Sorbus aucuparia L.), and goat willow (Salix caprea L.; see Konôpka et al. [31]). We found that the most intensive BATA_es was in goat willow (14.5%), and the least intensive was in common rowan (5.1%). BATA_es in common aspen was 8.7%. However, these measurements were performed in 2021. Therefore, they did not cover the winter season of 2021/2022. If we assume that bark browsing in the winter season of 2021/2022 for these three tree species was similar to that in 2019/2020 and 2020/2021, we can estimate the total BATA_es for the winter seasons of 2019/2020, 2020/2021, and 2021/2022 together. Specifically, the estimated three-winter-season values of BATA_es might be about 22% (calculated as [14.4/2] × 3), 13% ([8.7/2] × 3), and 8% ([5.1/2] × 3) for goat willow, common aspen, and common rowan, respectively. These estimates indicate that browsing intensity in goat willow might be comparable to that in Douglas fir and silver fir. On the other hand, BATA_es in common poplar might be about 2/3, and in common rowan, only about 1/3 of that in the fir species.

Our results showed that while bark browsing in distances up to 50 cm and over 175 cm was negligible, intensive damage was recorded between 76 and 150 cm from the ground level, with the maximum occurring at a height of 101–125 cm. These findings align with those reported for red deer in other regions and tree species. For instance, in Scotland, the majority of wounds on stems caused by red deer browsing were found between 51 and 150 cm [32]. Similarly, the highest percentage of stem bark browsing on common rowan was observed between 101 and 150 cm, with the upper occurrence at 200 cm [31] in Slovak forests. Another study from Germany [33] demonstrated that red deer typically stripped bark from the tree stem at heights between 80 and 170 cm, with an upper occurrence at 200 cm. Practically, these heights of maximum intensity of bark browsing correspond to the height of a red deer’s shoulder, which may vary slightly across regions [34].

Our further results also showed that stem thickness (expressed in three diameter classes) influenced BAB but not BATA. Logically, the larger the diameter class, the more bark surface, and consequently, a larger potential for browsing. Our previous study [31] conducted at the same site showed that, for rowan and aspen, the most intensively browsed trees were in the diameter class above 6 cm, and the least browsed occurred in the diameter class under 3 cm. However, these differences among the diameter classes were not observed for goat willow, where trees of all sizes were intensively browsed. We observed that because goat willow was the most attractive tree species, ruminating ungulates were willing to browse on very thin stems, which was complicated due to stem properties. Very thin stems are highly flexible, and browsing might be difficult because ruminating ungulates cannot press their teeth sufficiently against the stem surface. As for Douglas fir and silver fir, even trees in the diameter class under 35 mm were intensively browsed. This might indicate that these species are very attractive as forage and/or their stems are more rigid than in the common poplar and common rowan trees.

Our bark surface model, considering the stem sections situated 76–175 cm from the ground level, showed values ranging from 780 cm² (for a diameter d_1.3 of 20 mm) to 3330 cm² (when d_1.3 equals 85 mm). To provide a clearer picture, this represents a bark path in square dimensions of 28 × 28 cm and 58 × 58 cm for diameters of 20 mm and 85 mm, respectively. This represents the maximum bark surface available (forage potential) for red deer in a “comfortable” position. Our previous study [35] provided outputs for bark thickness and biomass (dry mass) in some broadleaves. The models allow the conversion of bark surface to biomass, enabling the calculation of the browsed bark quantity and the quantity of bark biomass available for browsing. Unfortunately, this kind of model is not available for Douglas fir and silver fir.

Our results clearly indicate that both Douglas fir and silver fir are “palatable” species and, therefore, extensively browsed by ruminating ungulates. Although there are some measures that can protect trees from browsing (e.g., fencing, repellents, protective caps placed on terminals, etc.), most of them are rather expensive and not 100% efficient. Therefore, regulating the population density of ruminating ungulates might be the best measure to reduce browsing damage to the palatable species of forest trees.

4.2. Bark Recovery

Regarding bark recovery (some authors use the term “healing”), RABA is a very suitable indicator to describe inter-species differences. In our case, silver fir shows statistically higher values (even though the differences were not very large) than Douglas fir. The RABA after 2–4 vegetation periods (2020, 2021, 2022, and 2023) was 62.5% in Douglas fir and 69.6% in silver fir. We could not find any studies showing the recovery rate (usually expressed in centimeters of recovered bark from the damaged edge of the wound) for Douglas fir and silver fir. However, the available literature focused on a variety of forest tree species (e.g., [36,37]) showed that the recovery rate is usually about 15 mm per year. This means, for instance, that for trees with a stem section diameter of 50 mm (equivalent to a circumference of circa 160 mm) and a stripping extent of 70% (i.e., the width of a wound of 112 mm), bark healing might theoretically take nearly four growing seasons. In fact, this theoretical calculation is not far from our results, as, after approximately three seasons (the average value for 2, 3, and 4 periods), wounds had recovered on about 2/3 of the stripped area.

Bark recovery after stripping is very likely influenced by a number of factors. The basic traits of the wounded trees (e.g., age and species) and the characteristics of wounds, such as severity, size, and location in which wounds occurred, determine the process of recovery [38]. Regarding silver fir, Pusz et al. [39] illustrated a variety of fungi associated with silver fir, noting that their composition changed across different stages of recovery. Moreover, they illustrated that recovered wounds predominantly contain Trichoderma atroviride P. Karst, a fungus antagonistic to many fungal pathogens. Hence, they assumed that this species can support the process of wound regeneration.

Delavaux et al. [40,41] demonstrated that recovery from bark harvesting for twelve medicinal tree species lasted at least two years and differed significantly among the species. Several other experiments have shown that the most important factor for successful recovery is the humidity of the wounded surface [42,43]. Delvaux et al. [40] conducted an experiment that removed bark only partially, leaving a thin bark layer with the cambium attached to the stem. They proved the protective effect of the remaining bark layer in promoting bark recovery. Moreover, Delvaux et al. [40] stated that trees severely attacked by insects or fungi pathogens had to manage their energy budget by compartmentalizing the wounded tissues to avoid the spreading of insects and pathogens within the stem rather than forming new wood tissue. Therefore, infections (fungi, insects, etc.) may considerably reduce the bark recovery rate in trees.

Regarding the physiological aspects of bark stripping, Zwieniecki et al. [44] stated that it interrupts the water relation between the bark and wood, affecting the water conduction between roots and foliage. As trees need large amounts of water, they need to develop mechanisms to protect against disruptions in their water balance and be able to rebuild the water pathway. Some studies have focused on the short-term responses to bark wounding (spanning from a few days up to a couple of months) expressed through compartmentalization, wound callus formation, wound reaction of the parenchyma cells, and lignin distribution in the xylem, among other factors [42,45,46]. Mwange et al. [43] noted that the initial steps in bark recovery are auxin-dependent. However, the first reaction that a tree initiates to protect itself after bark harvesting is the production of a callus directly in contact with the wound. This callus is formed from undifferentiated xylem cells at the stage of primary wall formation and consists of parenchymatous tissue without vessels, fibers, or any ray structures [42].

Nevertheless, knowledge of wound recovery processes in specific tree species concerning tree/stand and environmental conditions is urgently required. This demand is particularly relevant for the long-term perspective of tree survival and vitality, as well as future timber quality. It is essential to note that scientific attention is needed not only for bark stripping caused by ruminating ungulates (typical for young tree growth stages) but also for stem wounds resulting from harvest and skidding activities (frequent in middle- and old-aged forest stands, e.g., [47]). All these damages disrupt the concept of sustainable forest management; therefore, scientific findings can provide approaches for their reduction or mitigation, ultimately benefiting the forestry sector.

5. Conclusions

Our results illustrated the susceptibility of both tree species to bark browsing by ruminating ungulates, in our case, fallow deer, mouflon, and especially red deer. Despite the fact that the trees were exposed to bark browsing only within three winter seasons, all individuals were wounded. While inter-specific differences were significant, they were not large and may not be relevant from a practical forestry perspective. Therefore, we may suggest that Douglas fir (as an introduced species in Europe) is not a suitable substitute for “traditional” tree species under the conditions typical of most European countries. It is very probable that further mass plantation of this species does not support the intentions of sustainable forestry, aiming to fulfill all socio–economic roles of forests in these countries.

We can assume that, even though the recovery rate was rather high in both species (full recovery in young trees may last for approximately four growing seasons), the potential infection of stems by fungal diseases or wood deformations must not be underestimated. However, we did not research these aspects and cannot draw any well-grounded conclusions from them. Very likely, these kinds of harm can be considered as serious, with both ecological and economic consequences. Therefore, measures in both forest management and game management must be taken to ensure the sustainable development of forest stands. These measures, especially the protection of trees from browsing (group protection by fencing and individual protection by chemical or mechanical tools in forestry and the reduction of ruminating ungulate populations in game management), are crucial. Considering a broad perspective, the harmonization of interests among different stakeholder groups (i.e., foresters, hunters, and farmers) is currently an important aspect of sustainable development in the forestry, agroforestry, and agriculture sectors.