A Decade of Sanitary Fellings Followed by Climate Extremes in Croatian Managed Forests

Abstract

Forests in Croatia are characterized by higher levels of biodiversity in species composition. Three significant events occurred in Croatian forests over the past ten years, all of which have a common denominator—sanitary felling. The challenge in the sustainable development of forests started with the ice storm of 2014 that amounted to damage and raised costs in forest stands to EUR 231,180,921. The second challenge was in 2017 when the bark beetle outbreak occurred in the Gorski Kotar region. In December 2017, a windstorm in the same area caused damage to approximately 500,000 m³ of wood stock. The third climate extreme was in the summer of 2023 when three storms with strong winds and heavy rain damaged even-aged forests of common beech and pedunculated oak. The damage was substantial: 3,954,181 m³ of timber was mostly broken and destroyed across 21,888.61 ha of area, and the most damage was in the pedunculate oak forests of Slavonia, i.e., Quercus robur subsp. Slavonica, at 1,939,175 m³. For the main meteorological stations in lowland Croatia, data on precipitation amounts (mm) and wind speeds (m/s) were collected for the period 1981–2023, and the results of our analysis for the last decade are presented. Meteorological drought was analyzed using the rain anomaly index RAI. Data regarding open space fires in the Mediterranean karst area of Croatia were collected from the Croatian Firefighting Association, and the calculation of the burned area index (BAI) was determined. Throughout the entire area of Gorski Kotar County, a sample of permanent plots was set and used to assess the extent of forest damage from the ice storm in 2014 and for the establishment of permanent monitoring of the recovery of trees and forests damaged by the ice storm. The monitoring of bark beetles in the Gorski Kotar region started in 1995 and is still in progress. The aftermath of bark beetle outbreaks in two uneven-aged silver fir stands was studied after a bark beetle outbreak and a sanitary felling of 4655.34 m³. In the area of lowland Croatia, a statistically significant and positive correlation was found between sanitary fellings, maximum wind speeds, and rain anomaly indices in even-aged forests. In conclusion, sustainable development will be at risk due to difficult recovery, rising costs, and overall climate change in the years to come.

1. Introduction

The conservation and sustainable development of forests has been the focus of scientists for many years. The sustainability of forests as reservoirs of biodiversity, economic development, carbon sequestration, and much more is in the interest of both national and global policy makers. Unfortunately, climate extremes have been affecting forests rapidly and frequently throughout Europe in the past decade [1,2]. Strong winds, followed by insect outbreaks or devastating forest fires, have left heavy ecosystem side effects and disruptions of abiotic conditions [3,4,5,6,7].

The impacts of climate change on forest ecosystems can be direct and indirect [8,9]. Direct impacts include the consequences of gradual changes in climate parameters and/or extreme weather events, such as changes in temperature patterns, heat waves, droughts, frost, storm winds, etc. Indirectly, climate change affects the productivity and functioning of forest ecosystems through changes it causes in other components of the ecosystem, such as pollinators, pests, disease vectors, and invasive species. According to Lindner et al. [10], the impacts of climate change on forest ecosystems can be both positive and negative. In northwestern Europe, a positive effect on tree growth is expected as a result of higher CO₂ concentrations, higher temperatures, and a longer growing season [11]. On the other hand, an increase in drought and the frequency of weather extremes will bring numerous negative consequences, with the negative effects likely to outweigh the positive ones. In any case, the assumption is that the impact of climate change will be significantly reflected in geographically marginal populations [12], in ecosystems that are weakened and destabilized by anthropogenic activity [13], or in ecosystems that have a very narrow ecological niche.

Strong wind can be the catalyst of bark beetle outbreaks, especially in European spruce (Picea abies (L.) Karsten) and silver fir forests [14,15,16]. Økland and Berryman [15] and Wermelinger [16] reported that bark beetle outbreaks typically develop from one to three years after a disaster. The importance of bark beetles in silver fir and European spruce stands is a significant factor influencing the increase in biomass quantity and the reduction in its quality. Ips typographus (L.) is the most dangerous insect in European coniferous forests [17]. During the second half of the twentieth century, during a period of 50 years (1950–2000), the gradations of I.typographus influenced the dieback of about 150 million cubic meters of European spruce, following adverse abiotic factors [18,19]. Long-lasting periods of droughts or weather-related events that occasionally occur in time and space may cause bark beetle outbreaks as a result of weakened defensive mechanisms in spruce or because the bark beetle population increases up to that level when they become primary pests and can successfully colonize and kill a fresh and healthy tree [20]. Genetic and ecological mechanisms have facilitated adaptation to changes throughout history and have resulted in the development of optimal ecotypes adapted to local conditions through natural selection. The best examples are species from the genus Quercus sp. [21]. However, at the same time, in conditions of sudden climate change, the ecological optimum changes, but with a certain time lag before the optimal ecotype adapts to the newly created conditions [22] (Bürger and Krall, 2004). In close-to-nature forest management [23,24] an improvement in local environmental conditions, for example, the higher amount of deadwood volume, can be beneficial to the abundance of species with particular traits, while other species may decline. The intensification of natural disasters recorded in the last decades of the 20th century raises great concern regarding the disruption of the economic and ecological functions of forests [25]. Other authors consider ice storms as the most severe disturbance agents in forests of Europe [2,5,26]. In the past two decades, many authors have reported that ice storms influenced Western European countries most dramatically and greatly damaged Central and Eastern Europe [2]. The damage of ice storms to the forest ecosystem varies [27], from broken branches in individual tree crowns to catastrophic damage to entire forest stands, forest traffic infrastructure, and similar. Such a variety in damage affects the possibility of a quick recovery for both forest structures and forest management levels [28,29].

Wildfire behavior has changed across many different regions and forest types in recent decades, with negative effects on forest ecosystems [30,31]. Some authors claim that fire is a natural element in Mediterranean ecosystems [32,33], which needs to be considered when analyzing the impacts of fire on vegetation. However, forest fires and wildfires significantly disturb terrestrial ecosystems [34]. In the Mediterranean Basin, also in Adriatic climates, warming could amplify their severity and increase their frequency, negatively impacting forest ecosystems [35,36]. It is often highlighted that forest disturbances are natural processes and activators of system dynamics, often facilitating biodiversity [37,38,39,40], and are promoted in a way as an independent adaptation of forests to changing environmental conditions [38,39,41,42]. Deadwood in forests comes in different forms (e.g., snags or coarse woody debris) and decay stages. Depending on these forms and the stand itself, biodiversity can be enhanced, for example, as in the case of saproxylic beetles or woodpeckers [40,43,44,45,46]. Still, disturbances in forest ecosystems are predominantly perceived as negative due to various negative consequences, affecting the economy by increasing costs, decreasing revenues, and changing forest structures, species composition, and resource availability [28,29,47,48,49,50,51,52]. Deviation in stand conditions leads to uneven tree distribution, forest gaps, and disturbed relationships in the soil conditions, which can lead to the disappearance of the upper soil layer and reduction in biological activity in the soil, which can become unfit for the acceptance and success of tree species. The analysis of natural disaster trends has shown that damage from all causes (wind, fire, bark beetles, and other biotic and abiotic factors) has increased significantly from 1950 to 2019 across Europe [53]. The estimated average annual damage in Europe is 52.4 million m³ of wood, calculated for the period from 1950 to 2019. However, if only the last 20 years are considered, this average has risen to around 80 million m³ per year, which is equivalent to 16% of the total annual cut in EU forests. At regional and national levels, damage caused by natural disasters sometimes even exceeds 100% of the planned cut, which ultimately calls into question sustainable forest management.

The reasons for the sanitary felling of trees are economic and practical [54]. The amount and frequency of sanitary felling can be used as an indicator of forest health. The ratio of annual sanitary felling to total annual tree felling (allowable cut) can indicate whether the ecological stability of forests has deteriorated [55]. Sanitary felling is widespread and is carried out in temperate forests [56]. Sanitary felling is a forest operation aimed at ensuring economic returns after a natural or anthropogenic disturbance that caused damage to and mortality of forest trees. In addition to economic reasons, there are other justifications for felling damaged forests. These include reducing the fuel available for fires, reducing the risk of pests and pathogens from dead and dying trees, safety and security issues due to standing dead trees, and aesthetic reasons related to the appearance of damaged forests [57]. In the future, extreme weather conditions and events will increase in frequency and intensity, with high temperatures and summer droughts across Europe [58] and strong winds in central and Western Europe [59,60]. In Croatian forestry, conifer stands (silver fir and European spruce) are managed as uneven-aged stands, which in an ecological sense, has many benefits over even-aged stands (such as oak, ash, and beech). These benefits include increased resistance to defoliators, sap-suckers, and stem and root feeders. In the aftermath of bark beetle outbreaks, the uneven-aged structure of conifer stands is lost, and the forest regeneration phase is the most vulnerable stage to insect damage. In Europe, the major threat to seedling survival in coniferous forests, generally of an even-aged structure, is the pine weevil, Hylobius abietis (L.). Without preventive measures, H. abietis causes high levels of seedling dieback in European forestry and disrupts natural forest regeneration [61]. In Croatian forestry, seedlings are not protected by prophylactic treatment with insecticides such as synthetic pyrethroids and neonicotinoids, as their use is prohibited in forests and on forest soil because of FSC (Forest Stewardship Council) certification, which promotes sustainable forestry. In the case of uneven-aged conifer stands, standard practices in Croatian forestry, such as associational effects, lessen the damage caused by insect pests through different mechanisms.

The aim of this research is to present climate extreme events such as meteorological drought, ice storms, and hurricane winds, and their connection to sanitary felling over the last decade.

2. Materials and Methods

For the main meteorological stations in lowland Croatia, Gradište (Lat: 45.15; Lon: 18.7), Zagreb-Maksimir (Lat: 45.81; Lon: 16.03), and Gospić (Lat: 44.55, Lon: 15.36) and Delnice (Lat: 45.4, Lon: 14.8), both in Dinaric Croatia, data on precipitation amounts (mm) and wind speeds (m/s) were collected for the period 1981–2023, and the results of this analysis for the last decade are presented. Meteorological drought was analyzed using the rain anomaly index RAI [62] according to the following equation:

$$RAI=\pm 3={\displaystyle \frac{P-p}{E-p}}$$

where P—is the precipitation (mm); p—is the mean precipitation of a certain period (mm); and E—is the mean value of the 10 highest recorded monthly precipitation amounts (mm).

Raziei [63] stated that the application of the rain anomaly index (RAI) can represent a satisfactory alternative to more complex drought indices, such as the standardized precipitation index (SPI). Rain anomaly index (RAI) values above 2 are characterized as very dry years, and below −2 as very humid years [64,65]. Data regarding weather conditions (sessional, annual, maximal and minimal temperatures, and precipitation) were gained from the Croatian Meteorological and Hydrological Service. The connection between climate extremes and sanitary felling was analyzed using Pearson’s correlation. All data were processed in the Statistica v. 14.0.0.15., San Ramon, United States [66].

Data regarding open space fires in the Mediterranean karst area in Croatia were collected from the Croatian Firefighting Association and the calculation of the burned area index (BAI) was determined according to the following equation:

$$BAI={\displaystyle \frac{area affected by fire}{number of fires}}$$

Throughout the entire area of Gorski Kotar County, there was a sample of permanent plots from the first national forest inventory conducted between 2007 and 2009 [67]. The same sample of plots was used to assess the extent of forest damage from the ice storm and for the establishment of permanent monitoring of the recovery of the trees and forests damaged by the ice storm, as well as other biotic and abiotic factors. We selected the area affected by the ice storm based on the research by Šimić Milas et al. [68], in addition to the area satisfying the following conditions: (i) an area affected by the ice storm and (ii) beech-fir forests. In the area significantly affected by the ice storm, there were 122 plots of the national forest inventory, within which 20 plots were randomly selected for detailed measurement and assessment, and for the establishment of recovery monitoring. During the summer of 2015, a repeated measurement and assessment of these 20 plots was carried out according to the manual for the implementation of the national inventory [68] and an additional assessment of ice storm damage to individual trees on the plot. During the first national inventory, the plots were positioned using a metal wedge and GPS coordinates, which were used to find the center of the plots using a GPS device, and the wedge itself was found with a metal detector. All the trees on the plot with a radius of 20 m with a diameter breast height greater than 10 cm were measured. For each tree, the position and condition of the measurement at the first inventory were reconstructed. This included defining the variables that will be measured further (position, diameter, height, etc.) and whether the tree is alive or dead. Tree damage was assessed through five variables: (1) alive–dead, (2) crown damage, (3) tree inclination, (4) top damage, and (5) trunk damage. Crown damage is a typical form of ice storm damage and is a good indicator for assessing the possibility of tree recovery. It refers to the percentage (in 10% increments) of broken branches and twigs in the crown. Subsequently, by unification, narrower damage categories were formed: (A) destroyed crown (classes 70, 80, 90, and 100%); (B) significantly damaged crown (classes 40, 50, and 60%); (C) moderately damaged crown (classes 20 and 30%); and (D) undamaged crown (classes 0 and 10%). The tree inclination is the angle between the current tree axis and the ideal (vertical) axis, and this inclination is the result of the crown being overloaded by ice and wind. The inclination categories were upright; bent less than 45°; bent more than 45°; a crown that touches the ground; and an uprooted tree. An uprooted tree is a lying tree with its roots torn out. For every tree, especially young trees, the vegetative top is important. The status of the tree top refers to determining the fracture of the tree top itself. The trunk damage is described through three types of damage: undamaged, vertical cracks, or snapped trunk. Due to the heavy weight of the ice and strong winds, and with the tree’s strong rooting, the trunk can break from ground level to a higher level (snapped category), but also in longitudinal splitting due to the breaking off of larger branches or the rotation of the crown (vertical cracks category). Habitat variables that significantly influence tree rooting and stability were taken from the first national inventory. By combining the above-estimated variables for five groups of different tree damage, a complex, comprehensive variable, the tree damage coefficient was derived (KKS). Detailed information regarding this methodology can be found in previously published works by Teslak et al. [69] and Ficko et al. [70]. Based on the measurements, further structural variables of the stands were derived, such as stand basal area, growing stock, canopy, mean DBH, mean tree height, slenderness index, quantity and shares of damaged and destroyed growing stocks, etc. A second, repeated measurement was made in 2020, with an additional increment core taken from the trees to assess the recovery of tree growth. The second inventory process was identical to the 2015 measurement and assessment, with the addition of a core taken from the tree. In 2020, one short (up to 10 cm) increment core per tree was taken from 156 European beech (Fagus sylvatica L.), 85 silver fir (Abies alba Mill.), and 43 sycamore maple (Acer pseudoplatanus L.) trees at a height of 1.30 m from the east side of each sample tree. We sampled up to 10 trees per sample plot that were closest to the sample plot center. Tree-ring series were dated following standard procedures, using visual and statistical crosschecking (Figure 1).

To compare the radial increment before and after the ice break in individual trees, we calculated the radial increment index, (R.G. i3/3):

$$R.G.{\displaystyle \frac{i3}{3}}={\displaystyle \frac{r3A}{r3B}}$$

where r3A is the radial increment three years after the ice break, and r3B is the radial increment three years before the ice break (Figure 1).

We analyzed the rate of damage to trees and stands as changes in stand structure five years after the ice storm and following the ice break, as well as the level of tree recovery through the dynamics of radial increment. Analysis of variance (ANOVA) was used to compare the radial increment coefficient with tree damage classes by tree species, with a significance level of 0.05. Statistical analyses and graphical representations of the results were performed in the Statistica v. 14.0.0.15., San Ramon, United States [66].

The monitoring of bark beetles in the Gorski Kotar region started in 1995 and is still in progress. The monitoring of the spruce bark beetle (I. typographus) and the six-toothed spruce bark beetle (Pityogenes chalcographus L.) was carried out using Theysohn-type barrier traps (wet traps) in which the bark beetles are lured with Kombiwit Tube pheromone preparations on more than 1200 ha of forests. To determine the difference in the share of sanitary fellings in the allowable cut in different years, ANOVA was used. Silver fir is the most common coniferous tree species in Croatia, comprising 7% of all tree species. Forests of fir and hardfern (Blechnum spicant (L.) Roth) can mostly be found in the Gorski Kotar region. The aftermath of bark beetle outbreaks in two uneven-aged silver fir stands was studied in two adjacent areas in July 2023 in the Gorski Kotar region, which included a total area of 12.031 ha at locations Sunger 1 and Sunger 2 (Lat: 45.20; Lon: 14.4), where sanitary felling was performed in 2017 and 4655.34 m³ of timber was cut. This part of the research was concentrated on the determination of tree species naturally grown in the area after performed salvage logging, as well as young forest health status, together with the determination of forest pests. Two transects per area were made, and the health condition of all woody plants was determined. The transect plots were ten meters wide and were marked with an RTK (Real Time Kinematic) GNSS (Global Navigation Satellite System) STONEX S900A terrestrial receiver. Statistical data processing was performed using the program Statistica v. 14.0.0.15., San Ramon, United States [66]. Factorial ANOVA—a univariate test of significance for plant type, healthy status, and their interaction—followed by the Tukey post hoc test was used. Factorial ANOVA (least square means) for the interaction between the plant type (conifers–deciduous) and their healthy status (damaged–healthy) followed.

3. Study Area

According to the valid National Forest Management Plan, it is estimated that the growing stock in the Republic of Croatia is 398 million m³, of which 302 million m³ is in state forests managed by the company Croatian Forests Ltd. and slightly over 78 million m³ is in private forests, while 17 million m³ of state forests are used by other legal entities [71]. According to the FSC Organisation [72], nearly three-quarters of the total forest area in Croatia is FSC-certified, which leads to the conclusion that Croatia has the highest share of FSC-certified forest area in the world, amounting to 2.8 M ha and 404 certified companies. Forests (high and coppice) cover 1.8 M ha (i.e., 63%), and shrubs, maquis, and other kinds of vegetation cover 0.7 M ha (i.e., 27%), while 0.3 M ha (i.e., 10%) are bare forest lands. According to the EU Biodiversity site [73], 38.1% of Croatia’s terrestrial territory is designated as protected areas, significantly above the EU value of 26.4%, with the target of 30% to be reached at the EU level by 2030, as set by the EU Biodiversity Strategy. The European Environment Agency further states [74] that the Natura 2000 network is a network of protected areas covering Europe’s most valuable and threatened species and habitats, thus being the largest coordinated network of protected areas in the world, extending across all 27 EU Member States. The ecological network Natura 2000 in the Republic of Croatia covers 36.67% of land and 16.39% of the sea along the coastline. The management of forests in Croatia is mainly performed by the company Croatian Forests Ltd., thus being the most significant state company that manages 75% of all forest areas in the country, of which 382,535.70 ha (68.38%) are located within the ecological network Natura 2000, balancing between the conservation status of species and habitats protected under the EU Birds and Habitats Directives and bio-economy.

4. Results

Meteorological drought in a ten-year period was analyzed using the rain anomaly index, RAI. Figure 2 shows rain anomaly indices for the last decade for meteorological stations in lowland and Dinaric Croatia. In lowland Croatia, three years were moderately dry, while there were no very dry or extremely dry years, while in the Dinaric area, two years were extremely dry.

Figure 3 shows the annual maximum wind speeds (m/s). The eastern part of lowland Croatia had a total of five events, with hurricane-force wind speeds causing damage to forest ecosystems, and Dinaric Highland Croatia had two such events.

Rainy years, in combination with hurricane winds, are favorable weather events for stand damage, such as windbreaks and windthrows (

Table 1. Years with climatic conditions for wind damage (windstorm and windthrow).

Region	Climatic Conditions		Damage
	Very Humid	Hurricane Wind	Windstorm	Windthrow
Lowland	2014; 2018; 2023	2014; 2015; 2018; 2019; 2023	2014; 2015; 2018; 2019; 2023	2014; 2018; 2023
Dinaric	2014; 2023	2020; 2023	2020; 2023	2023

). In the lowland Croatian area, at the Gradište meteorological station, such years were 2014 and 2018, and at the Zagreb–Maksimir meteorological station, 2014 and 2023. In Dinaric Croatia, at the Gospić station, it was the year 2023. For the Zagreb–Maksimir meteorological station, a statistically significant and positive correlation was found between the rain anomaly index (RAI) and maximum wind speed (r = 0.45 *, p = 0.036).

Table 2. Descriptive statistics of the percentage ratio of sanitary fellings to planned fellings.

	N	Mean	Minimum	Maximum	Std. Dev.
Vinkovci (lowland)	21	25.857	9.000	65.000	12.709
Zagreb (lowland)	21	22.000	16.000	47.000	6.411
Bjelovar (lowland)	21	13.905	8.000	31.000	5.576
Delnice (Dinaric)	21	28.476	7.000	82.000	20.949

shows the descriptive statistics of the percentage ratio of sanitary fellings to planned fellings for forest administrations in lowland and Dinaric Croatia. The maximum amounts of sanitary fellings ranged from 31% to as much as 82% in the area of selected forests in the Dinaric region.

In the area of lowland Croatia, a statistically significant and positive correlation was found between sanitary fellings, maximum wind speeds, and rain anomaly indices in even-aged forests, while this is not the case in the area of selected forests in Dinaric Croatia (

Table 3. Pearson’s correlation of percentage of sanitary felling with rain anomaly index and maximum wind speeds (* significant p < 0.05).

Region	Coefficient r/p Value
	Max. Wind Speed	RAI
Sanitry fellings—lowland	0.49 */0.022	0.55 */0.009
Sanitry fellings—Dinaric	0.27/0.22	0.06/0.78

).

In the fires observed in the period of 1 January–31 December 2023 (

Table 4. Open space fires in the Mediterranean karst area in Croatia (data source: Croatian Firefighting Association).

Open space fires in the Mediterranean karst area in Croatia											2023
Observed period	2018		2019		2020		2021		2022
	No. of fires	Burned area (ha)	No. of fires	Burned area (ha)	No. of fires	Burned area (ha)	No. of fires	Burned area (ha)	No. of fires	Burned area (ha)	No. of fires	Burned area (ha)
1 January–31 December	1875	3891	3850	19,129	3906	35,168	3250	14,707	4247	30,039	1951	3551
1 June–31 October	1330	3160	1090	2643	1063	1695	1577	8444	1607	11,628	928	2690
Observed period	Five-year average 2018–2022										2023
	No. of fires		Burned area (ha)			Average fire duration			BAI (ha/fire)		BAI (ha/fire)
1 January–31 December	3426		20,587			2 h 40 min			6.01		1.82
1 June–31 October	1333		5514			2 h 58 min			4.14		2.90
Observed period	2023/Five-year average 2018–2022
	Ratio of the No. of fires		Ratio of burned area			Ratio of the burned area index (BAI)
1 January–31 December	−43.05%		−82.71%			−69.71%
1 June–31 October	−30.40%		−51.22%			−29.90%

), the total burned area was estimated at 3551 ha, which is 82.71% less compared to the observed previous five-year average, while the burned area index (BAI) is 69.71% lower. If we look only at the period of the fire season (1 June–31 October 2023), 928 vegetation fires were recorded, which is a decrease of 30.40% compared to the previous five-year average, while the estimated burned area was 2690 ha, which is a decrease of 51.22%. These values resulted in a 29.90% reduction in the BAI.

Comparing 2023 with the ten-year average (Figure 4 and Figure 5) for the coastal and karst Mediterranean areas, a decrease in the number of fires by 34.99% and the estimated burned area by a large 85.24% was recorded, with an exceptional decrease in the BAI by 77.30%.

The reduction of all observed values compared to the five-year average is all the more significant because the five-year average no longer includes the observed values of vegetation fires from 2017, which in all observed indicators was the most demanding fire season in more than 20 years.

4.1. Ice Storm in 2014

In 2014, an ice storm of the most immense intensity and spatial extent ever recorded occurred in Slovenia and Croatia [75]. In terms of extent and total wood volume lost, this event was then the region’s most catastrophic natural disturbance on record. The storm created conditions for supercooled freezing rain from 30 January to 5 February. In the most severely affected areas, more than 200 mm of precipitation fell, which created a glaze up to 8 cm thick, while much of the central region of the storm experienced 100 mm of precipitation with ice accumulation around 4–5 cm [76,77].

According to the data from the Croatian Forests Ltd., in cooperation with the Faculty of Forestry and Wood Technology University of Zagreb and the Croatian Forestry Institute, estimates of damage caused during the ice-storm in Gorski Kotar region on 1 and 2 February 2014 (Figure 6) amounted to EUR 231,180,921, which includes damages in both state and private forests, together with damage caused by floods and torrents resulting from the ice melting, as well as damage to the traffic infrastructure network. Loss on the growing stock amounted to an estimated EUR 161,121,554, while damage to traffic infrastructure amounted to EUR 3,190,537. The same source stated that the recovery would be long-lasting and challenging and that costs would rise to an additional EUR 66,868,830.

In the 2015 survey, 1389 trees were recorded on the plots, of which 826 were common beech, 268 were silver fir, 169 were maple, and 126 were European spruce (

Table 5. Change in the number of trees separated by tree species in two consecutive inventories in the area affected by the ice storm.

	2015 Year of Measurement	Tree Felling	Share	Ingrowth	Share	2020 Year of Measurement
Tree Species	No.		%	No.	%	No.
Beech	826	146	17.7	74	9.0	754
Silver fir	268	56	20.9	15	5.6	227
Maple	169	40	23.7	14	8.3	143
E. spruce	126	49	38.9	4	3.2	81
TOTAL	1389	291	21.0	107	7.7	1205

). By 2020, 291 trees, or about 21% of the trees, had been cut through sanitary felling (Table 5). A total of 107 new trees, or about 7.7%, had grown into the sample so that in 2020, 1205 trees were recorded, which represents a decrease of 184 trees or about 13% (Table 5), representing changes in the structure directly caused by the ice storm.

Changes in the structure of stands damaged by ice break are a consequence of tree mortality or sanitary felling. The mean diameter at breast height of both broadleaf and conifer trees has increased due to significant damage to thinner trees. The same is true of tree height. This is particularly important in the context of preserving the selective multidimensional structure of beech-fir forests.

The number of trees per hectare of broadleaf trees decreased by almost 19% (82 trees), while the number of conifers decreased significantly less (about 4% or 5 trees per hectare). This data suggests that conifers are less susceptible and more resistant to ice storms and snow. The results on the reduction in the number of trees are accompanied by the results on the reduction in the stand basal area and growing stock, with the reduction in the growing stock of conifers being at a greater percentage because it is the larger trees that are most affected (about 8% of the growing stock of conifers was reduced). Interestingly, there was no significant change in the ratio (representation) of tree species to the growing stock (

Table 6. Basic structural elements (average diameter breast height, dA, average total tree height, hA, number of trees, N, stand basal area, B.A., growing stock, V, and proportion of species in growing stock, OS_V) of stands, according to two consecutive measurements and their change.

	Broadleaf (beech, maple)
	plots	dA	hA	N	B.A.	V	OS_V
	n	cm	m	ha−1	m2ha−1	m3ha−1	%
Ice storm area 2015	20.00	21.60	15.90	436.90	19.10	209.30	46.60
Ice storm area 2020	20.00	24.80	19.10	354.90	17.10	193.60	46.70
Change		3.20	3.20	−82.00	−2.00	−15.70	0.10
	Conifers (silver fir, European spruce)
Ice storm area 2015	20.00	34.40	21.29	129.50	16.80	239.90	53.40
Ice storm area 2020	20.00	40.10	25.60	124.50	15.70	221.40	53.30
Change		5.70	4.31	−5.00	−1.10	−18.50	−0.10
	All species
Ice storm area 2015	20.00	25.30	17.40	566.50	35.90	449.20
Ice storm area 2020	20.00	29.50	19.10	479.50	32.80	415.00
Change		4.20	1.70	−87.00	−3.10	−34.20

).

The results of the analysis of variance (

Table 7. Two-way ANOVA—multivariate test of significance for tree species’ (VR) crown damage categories (OK_kat), their interactions, and their radial increment coefficients (ir3/3).

Effect		ir3/3	ir3/3	ir3/3	ir3/3
	df a	SS b	MS c	F	p
Intercept	1	150.040	150.040	381.404	0.000 ***
VR	2	3.265	1.633	4.150	0.017 *
OK_kat_2015	3	6.062	2.021	5.137	0.002 *
VR × OKkat_2015	6	5.267	0.878	2.232	0.040 *
Error	277	108.968	0.393
Total	288	129.219

Statistically significant differences are denoted by asterisks as follows: * p < 0.05; *** p < 0.0001. ^a—degrees of freedom. ^b—sum of squares. ^c—mean square value.

, Figure 7) on the radial increment coefficient (tree recovery) in relation to the categories of crown damage assessed immediately after the ice break (2015) are not unambiguous. It is to be expected that the greater damage shows a lower recovery, but the results indicate that younger, thinner trees, especially fir, but also other species, even with severe damage, show successful recovery (coefficient ir3/3 greater than one).

This is related to the level of loss (opening) of stand structure and the opening of resources for previously suppressed trees. For the other crown damage categories (B., C., and D.), although there is no statistically significant difference, the lower the crown damage, the better the tree’s recovery, as measured by radial increment recovery (Figure 7).

Two consecutive (five-year interval) assessments of tree damage allow us to show the degree of damage to forests from ice storms, the level of recovery achieved through sanitary felling, and the progress in the tree recovery process. i.e., through the analysis of the condition of trees and stands six years after the ice storm. The extent and severity of the damage are indicated by the average damage coefficient (KSS) of almost 0.5 for broadleaf trees and 0.4 for conifers. Almost 34% of the broadleaf growing stock was destroyed (it cannot be recovered and needs to be removed), and an additional 24% was seriously damaged. Conifers were less affected, with up to 10% of the growing stock destroyed and an additional 8% of the growing stock seriously damaged (

Table 8. Indicators of the degree of ice storm damage and the level of damage repair (recovery) (habitat quality (BS), tree damage index (KSS), slenderness coefficient (KV), volume of destroyed growing stock (Vdes.), and volume of seriously damaged growing stock (Vsde.), the percentage share of destroyed growing stock (pdes.) and the percentage share of seriously damaged growing stock (psde.)) of stands, according to two consecutive measurements and their changes.

	Plots	BS	KSS	KV	Vdes.	Vsde.	pdes.	psde.
	Number				m3ha−1	m3ha−1	%	%
species	Broadleaf (beech, maple)
Ice storm area 2015	20	3.0	0.49	0.81	70.9	49.9	33.8	23.8
Ice storm area 2020	20	3.0	0.33	0.86	38.7	71.8	20.0	37.1
Change		0.0	−0.16	0.05	−32.2	21.9	−13.8	13.3
species	Conifers (silver fir, European spruce)
Ice storm area 2015	20	2.5	0.40	0.66	23.2	18.9	9.7	7.9
Ice storm area 2020	20	2.5	0.24	0.61	18.3	66.5	8.3	30.0
Change		0.0	−0.16	−0.05	−4.9	47.6	−1.4	22.1
species	All species
Ice storm area 2015	20	3.0	0.46	0.76	94.0	68.8	20.9	15.3
Ice storm area 2020	20	3.0	0.41	0.69	57.0	138.4	13.7	33.3
Change		0.0	−0.05	−0.07	−37.0	69.6	−7.2	18.0

).

Sanitary fellings were carried out, but only partially. This is indicated by the fact that there is still a lot of destroyed growing stock in the forest (20% of broadleaf and 8% of conifers). The total mortality coefficient of broadleaf and conifer trees (KSS) decreased by only 0.16. The long-term recovery process is indicated by the existence of a large proportion of seriously damaged growing stock, even six years after the ice storm (37% for broadleaf trees and 30% for conifers) (Table 4). At the same time, we cannot rule out that there was no new damage during the period due to windbreaks, droughts, plant diseases, and pests that were recorded in the 2020 assessment. This especially applies to damage to European spruce and silver fir from drought and, secondarily, from bark beetles.

The ice storms in the Gorsky Kotar region very likely influenced the increase in the bark beetle population, which subsequently led to the proclamation of a natural disaster in 2017. This was not only because of the bark beetle outbreak but also because these secondary pests, mainly I. typographus, attacked surrounding forests not initially affected by the ice storm.

4.2. Bark Beetle Outbreak in 2017

The monitoring of bark beetles in the Gorski Kotar region (total area of 21,300 ha) started in 1995 (Figure 8A) and is performed in 46 different compartments spread throughout the region, covering 1254.73 ha of the area.

The first peak was measured back in 2005 at 329,173 beetles, then a significant peak in 2017, just three years after the ice storm, with 654,587 beetles, and another peak in 2020, with 492,437 beetles. Over the 28 years of monitoring, most bark beetles were caught in June (Figure 8B).

The outbreak in 2017 resulted in the marking of 167.097 m³ of infested European spruce trees and, consequently, the cutting of 153.426 m³ of marked trees. Sanitary fellings were performed in 14 Forestry Offices (FOs) with different extents, depending on the infestation. Before the infestation, the volume of European spruce in the Gorski Kotar region accumulated to a volume of 2,279,876 m³ and ranged in share from 1.16 to 21.83% in different FOs in this region. In seven areas, a total of 58.125 ha of clear-cutting was performed due to disturbance management since clear-cuts are forbidden in Croatia. Most of I. typographus in the 28 years was caught in three FOs (Figure 8C)—Gerovo, Vrbovsko, and Tršće—with a total of 3,087,224 beetles.

In December 2017, a windstorm in the Gorski Kotar region caused damage to approximately 500.000 m³ of wood stock [78], with even more in silver fir and European spruce stands. Due to the difficult terrain conditions and the high amount of damaged and downed trees, not all damaged trees were removed from the forest stands in due time.

The analysis of sanitary fellings in the period from 2003 to 2023 showed a significant increase in the share of the allowable cut after the years with high climatic extremes in the Gorski Kotar region (Figure 9), especially after the bark beetle outbreak and the wind storm in 2017, which was followed by the ice storm in February 2014.

4.3. The Aftermath of Bark Beetle Outbreaks in Two Uneven-Aged Silver Fir Stands

The researched stand will, in the near future, go through the process of thinning, which is usually performed to increase the economic value of the stand and should prepare the new forest to be stronger and healthier, as well as less affected by pests. In the researched area along the transect plots, 2930 trees of woody plants were determined, with a total of 10 species. The most frequent species were rowan (Sorbus aucuparia L.) and European spruce (Picea abies). They were followed by silver fir (Abies alba), aspen (Populus sp.), and willow (Salix sp.) species. The health status of all the woody plants was recorded as follows: (1) gall–aphid, (2) gall–Diptera, (3) leaf gnawing, (4) leaf skeletonizing, (5) leaf–needle mines, (6) needle dieback, (7) tree dieback, (8) bark peeling, and (9) healthy status. Inspecting the transect plots revealed the frequent appearance of pests that attack the leaves of deciduous species. In the Sunger 1 plot, the most common species was rowan, with a share of 30.8%, the second was European spruce, with 30%, and the third was silver fir, with 15.7%. We determined that 73.7% of trees were healthy and without pests. The healthiest tree species was silver fir at 97.7%, followed by rowan at 96.4%. European spruce was without pests in 70.5% of the cases. The most common sign of pests was leaf gnawing (Figure 10A). It occurred on 16.2% of the trees, mostly on willows, aspen, beech, and birch (Betula pendula Roth). On beech, a great abundance of leaf gall was found, caused by beech gall midge (Mikiola fagi (Hartig. 1839)) in 82.5% of beech trees (Figure 10B). Most of the leaf gnawing and leaf mines on beech were conducted by beech-leafed miner beetle (Rhynchaenus fagi (Linnaeus. 1758)).

On the second transect plot of the researched area, Sunger 2, the most common species was European spruce with a 38.5% share, second were willows with 22.9%, the third was aspen with a 16.3% share, rowan was fourth with 14.3%, and silver fir was fifth with an 8.4% share. On the plot Sunger 2, 47.4% of the trees were healthy and without pests. Silver fir was the healthiest tree species with 94.1% healthy trees, followed by rowan with 93.1% healthy trees. European spruce was without pests in 82% of cases. The most common sign of pest was, again, leaf gnawing. It occurred on 43.3% of the trees, mostly on willows, aspen, beech, and birch in more than 90% of the trees for each species. On beech, again, a great abundance of leaf gall was found caused by beech gall midge (Mikiola fagi) in 85.7% of beech trees, as well as the presence of beech leaf miner beetle (Rhynchaenus fagi). In the coniferous species, the most abundant pest was the same as in Sunger 1, a spruce gall aphid (Sacchiphantes viridis (Ratzeburg 1843)) on European spruce (Figure 10C), which was found on 15.6% of spruces observed. Another pest was found on silver fir, a plant pathogen from the genus of ascomycetes fungi (Cytospora pinastri Fries) (Figure 10D).

We analyzed the differences in plant type (conifers–deciduous) and their healthy status (damaged–healthy) on the investigated plots. For a healthy status, plants with any damage (gall–aphid, gall–Diptera, leaf gnawing, leaf skeletonizing, leaf–needle mines, needle dieback, tree dieback, and bark peeling) were grouped. Only one plant of sycamore (Acer pseudoplatanus L.) and Scots pine (Pinus sylvestris L.) were found in one plot, so they were excluded from the analysis. First, there were no differences between the analyzed groups when considering a 0.95 confidential interval (Figure 11). However, factorial ANOVA showed statistical significance between variables and their interactions (

Table 9. Factorial ANOVA—univariate test of significance for plant type, healthy status, and their interaction (type × healthy status).

Univariate Tests of Significance
Effect	df a	MS b	F	p
TYPE	1	24,390.1	4.007	0.049 *
HEALTY STATUS	1	31,314.1	5.144	0.026 *
TYPE × HEALTY STATUS	1	28,275.5	4.645	0.035 *
Error	60	6086.4

Statistically significant differences are denoted by asterisks as follows: * p < 0.05; ^a—degrees of freedom. ^b—mean square value.

). Figure 12 shows that the differences in the plant types and their healthy status are visible. Nevertheless, the post hoc Tukey HSD test for these results did not indicate a statistically significant difference between the analyzed factors, which can be explained and affected by the large variability in the analyzed data (shown in Figure 11).

Figure 13 presents the interaction of two parameters—type × healthy status. Conifer plants tend to be healthier than deciduous plants on the investigated plots.

4.4. Summer Storms in 2023

In the summer of 2023, specifically on 19 and 21 July and 4 August, strong wind storms with heavy rain hit Slovenia, Croatia, Bosnia and Herzegovina, and Serbia. In the continental part of Croatia, from the west to the east of the country, it caused enormous material damage (Figure 8) and, unfortunately, also the loss of human lives. According to the Croatian Meteorological and Hydrological Service [79], preliminary data from automatic meteorological stations showed the most intense precipitation was recorded on the path of the thunderstorms in the area of the capital, Zagreb, where 34.6 mm was recorded in half an hour (from 16:00 to 16:30). Such an amount of precipitation, according to the assessment of extremes, can be expected on average once every 16 years. On a shorter scale, the 10 min maximum was 22.5 mm, which, according to the assessment of extremes, can be expected on average once in almost 50 years. The same source continues that although the recorded amount of precipitation does not represent an exceptional event, its short-term intensity combined with stormy winds caused enormous material and human damage. A wind gust of 114.8 km/h from the northwest direction was recorded at the Zagreb airport location at 4:10 p.m. For comparison, at a near location, Zagreb–Maksimir, one can expect to exceed the (average 10 min) wind speed of 40.5 km/h on average once in 20 years, or 42.8 km/h in 100 years. In the east of the country, at the measuring station Slavonski Brod, the wind recorded at 18.00 h was as much as 118.4 km/h. In contrast, at the Gradište location, the wind knocked down a pole with wind speed and direction sensors, so there was an interruption in the measurements. The preliminary results show that this may be also the absolute maximum of wind gusts for this location which should be further confirmed when the data passes all levels of control prescribed by the standards of the World Meteorological Organization.

Damage to forests due to these three storms was substantial. According to the data given by the national company Croatian Forests Ltd. Zagreb, Croatia, in total 3,954,181 m³ of timber was damaged (mostly broken and destroyed seldomly downed) on 21,888.61 ha in the area (Figure 14 and Figure 15), of which the most was in the pedunculate oak forests of Slavonia, i.e., Quercus robur subsp. slavonica (Gáyer) Mátyás at 1,939,175 m, followed by beech at 638,063 m³, sessile oak (Quercus petraea (Mattuschka) Liebl.) at 159,612 m³, silver fir and European spruce at 31,864 m³, and the rest of the tree species at 1,185,467 m³. The most damage was in the Forest Administration (FA) of Vinkovci in the east of the country, where in total 2,686,626 m³ of timber was damaged, of which 1,755,000 m³ was Slavonian pedunculated oak in Spačva basin, in an area of 16,606.54 ha. By the end of 2023, 700,000 m³ of timber was removed (cut, processed, and extracted) from the forest stands, with a plan for 2024 to remove an additional 1,400,000 m³ of timber (of which 1,100,000 m³ is in FA Vinkovci), including saving as much veneer logs of pedunculated and sessile oak as possible, since most damage was in the last age class (age from 120 to 140 years) of these even-aged forests. In 2025, the rest of the timber destroyed will be removed from the forests. A challenge for the next decade in the sanitation of these stands will be in (1) the afforestation of thousands of hectares of land; (2) the removal of damaged destroyed and downed trees and stems; (3) stand preparation; (4) the production of forest seedlings; (5) the planting of tree seedling; (6) the general lack of workforce, and (7) the lack of the acorn crop.

5. Discussion

Mixed forest stands are more resistant to biotic and abiotic disturbances than pure forest stands [80,81]. The main cause of large-scale natural disasters in European forests is wind. In the last 70 years, wind has caused about 46% of all forest damage compared to other natural disasters, with an annual average of about 24 million m³ of damaged wood biomass [53]. The extent of forest damage depends on the duration of the wind, the maximum wind speed, but also on the precipitation immediately before and during the event, the age of the forest, the type of tree, the management method, etc. Wind and drought damage has increased in temperate forests [82], followed by an increase in the number of dead trees, which leads to a strong impact on the forest ecosystem, ecosystem functioning, and ecosystem services [83,84]. In the last decade, mixed forest stands have suffered from extreme climatic events and biotic factors in Croatia. High rainfall causes instability in the root structure in such a way that trees are more susceptible to uprooting. Precipitation with stormy winds leads to a higher risk of tree uprooting [85]. In the absence of rainfall, tree roots usually withstand such wind gusts, but with increased wind speed, trees will break along the trunk rather than be uprooted [86]. According to Csilléry et al. [82], extreme rainfall leads to complete soil saturation with water, thus creating conditions that make trees more prone to wind-blown debris in storms. In Croatia, such years have occurred three times in the last decade: 2014, 2017, and 2023.

In lowland Croatia, sanitary fellings have been positively correlated with the rain anomaly index (RAI), showing that very wet years in which floods occur also affect tree mortality. Namely, our lowland forests are the largest natural retention areas [87], into which excess water from rivers is released, thus preventing floods. However, if this water is released into the forest during the vegetation period and if floods last for a long time, there is a lack of oxygen in the soil, increasing carbon dioxide concentration in the soil and root necrosis, leading to extraordinary tree mortality [88].

The amounts of sanitary fellings can be many times higher than the allowable (permitted) cut [89], all with the aim of enabling the rehabilitation of areas affected by disturbances and tree death. In the study area, sanitary fellings amounted to up to 82% of planned felling, which is a high percentage and indicates reduced vitality and significant disturbances in forest ecosystems. However, such fellings also bring certain consequences for forest ecosystems, i.e., economic and ecological problems. Economic problems relate to a decrease in wood stock, a decrease in biomass production, disrupted sustainable management, a lack of natural regeneration, and the weeding of forest stands, while ecological problems relate to a decrease in biodiversity, the disappearance of stands for certain plant and animal species, and a decrease in the value of generally beneficial forest functions.

Csilléry et al. [82], in some stands, found a positive interaction between drought and storm intensity. Drought weakens trees, and they become more prone to trunk breakage due to adverse wind action. In other stands, they found a negative interaction between drought and storm intensity, where excessive rain likely led to complete soil saturation with water, thus creating conditions that make trees more prone to becoming wind-blown in storms. In our research in lowland Croatia (Zagreb–Maksimir station), a positive and significant correlation was found between the drought index (RAI) and maximum wind speeds, showing that very humid conditions, i.e., high precipitation followed by storms, cause wind-blown trees.

The largest number of fires in the coastal and karst Mediterranean areas of Croatia in 2023 was recorded in the month of February, slightly less in March, and, as usual for the summer months, in July and August, which followed the characteristics of 2022 (except for February). February and March together had a share of 30.60% in the total sum of all vegetation fires, while July and August followed with a share of 25.89%. Fires in February and March point to the burning of agricultural land and weeds.

In some parts of Europe (e.g., the Dinaric Mountains and some areas in the Alps), ice storms are frequent and can occasionally cause extensive damage [75]. Considering the spatial scope, they can be local or regional, such as the last ice storm in Croatia in 2014. According to Pernar et al. [90], several conclusions regarding the ice storm in 2014 were made, as follows:

(1)

The influence of ice weight on the bending, tilting, and spreading of trees was more pronounced in beech (Fagus sylvatica L.) and other broadleaved trees, as opposed to conifers.

(2)

Broken and split trunks were represented in beech at an amount of 22.6% (other broadleaved species at 20.4%), which, in addition to fallen trees, represented an additional loss. In contrast, the loss of silver fir and European spruce trees due to this form of damage was only about 1.5%.

(3)

The largest proportion of trees with broken tree tops was represented in European spruce (39.7%), then other broadleaved species (39.2%) and beech (35.2%), with the least for silver fir (28.3%).

(4)

The timber volume of fallen and broken trees and trees with large canopy damage (over 80%) amounted to about 103,000 m³ or 63% of the prescribed 10-year allowable cut for 2012–2021. The damaged timber volume of beech, which participated in the mixture ratio, was 28.4% (almost doubled the amount of the prescribed 10-year allowable cut).

(5)

The damage made to forest roads was at a length of 753 km. The estimation of the recovery costs of lysis amounted to EUR 445,956.

(6)

The damage to wild game is observed through direct mortality of large and small game and through long-term changes in population dynamics caused by significant habitat changes. The mortality in game caused by storms is quite difficult to determine due to fallen timber, impassability, etc. Increased anthropogenic disturbance due to recovery operations leads to increased stress on the area’s wild animals. As a result of increased calamities, the direct measure is a ban on hunting until the population recovers to a satisfactory state, and the indirect measures are supplementary nutrition, reduced harassment, etc.

Vuletić et al. [91] concluded that the costs in the Gorski Kotar region regarding the 2014 ice storm are higher than first assumed and that the total forest damage amounts to EUR 942,252,183. The results showed that 56,021.68 ha of forests were damaged, of which 19,245.79 ha of forests have been seriously damaged while 9808.22 ha of forests have been destroyed. The total costs included the loss of timber volume in damaged and destroyed stands, the total estimated damage to forest roads, skid roads, and trails, the cost of recovery based on identified priority areas, and the loss of value for forest ecosystem services (FESs). Unfortunately, another challenge is that bark beetle outbreaks hit the same area only three years later. It is known that in managed forests, there are two main types of bark beetle controls to prevent the emergence of the beetle outbreak and to reduce the size of the beetle population [92]. Sanitary fellings, being the first one, involve the removal of downed trees and wood debris, as removal of available breeding material is recommended within two years after the climate extremes [93], along with continued sanitation felling and the removal of standing beetle-infested trees to avert the emergence of possible beetle offspring. Unfortunately, due to the high amount of damaged timber and low trained workforce, this is not always possible to be completed in the desired time frame.

Still, after considerable damage to forest stands caused by extreme weather conditions, coarse wood debris in windthrow gaps can become a niche for saproxylic insect diversity [94]. Croatian uneven-aged conifer forests are generally characterized by higher levels of biodiversity in terms of tree age and species composition and thus suffer lower damage caused by herbivores [95]. Bark beetle catches, after the outbreak in 2017, when the climax of the bark beetle population expanded all suitable breeding materials, are typical in the bark beetle outbreak aftermath. European experiences concur with average catches per trap ranging from 13,535 I. typographus per trap in Slovenia in 1998 [96,97] to 10,000 to 44,000 per trap from 1995 to 2000 in Sweden [98]. Regarding the average P. chalcographus catch per trap in Slovenia, as many as 581,222 individuals on average were caught (Slovenia Kranjsko polje) [99], although less than in Austria, with 773,300 bark beetles per trap [100].

Dobor et al. [1] stated that salvaging trees destroyed and damaged by wind and bark beetles is extensively applied in the coniferous forests of Europe. They continued, saying that new management responses to changing forest disturbance regimes are necessary, with a focus on forest resilience. On the other hand, Trotsiuk et al. [101] concluded that the response of forest productivity to climate extremes strongly depends on environmental and site conditions. Christiansen [102] concluded that even though many aspects of forest stand adaptation have been comprehensively researched, changes in the growth of different tree species, the impact of pathogens and insect-assisted migrations, stand health status, and knowledge of the resistance of forests to natural calamities are additional tools in the adaptation of forests and forestry in today’s changing environment.

There is a large amount of research that studies the effects of storms on terrestrial animals and birds [103,104,105,106]. High energy consumption, frequent injuries, and exposure to vibration, noise, gases, dust, and other hazards place operational forestry among the most physically demanding jobs [107]. Some studies suggest that tree felling of old (mature) trees reduces the stability of the stand from the perspective of minimizing some climate extremes, such as storms [108], while others [109]) highlight that the probability of ice damage decreases significantly with the size of the tree diameter (DBH), i.e., with larger trees being less susceptible to damage. On the other hand, after the storms and climatic extremes, positive changes in wild game can be manifested in the development of layers of ground growth and bushes due to the penetration of light into the lower layers of the stand, which leads to an increase in food potential and shelter opportunities.

6. Conclusions

Two basic attributes are particularly important for the implementation of emergency rehabilitation measures after natural disasters: the severity of the natural disaster and the size of the area affected. In the past ten years, changing and extreme weather conditions have affected forests repeatedly and severely in Croatia and Europe, bringing difficult recovery and rising costs and affecting sustainable development. In lowland Croatia, a significant correlation was found between the amount of sanitary felling and climatic extremes, such as floods and hurricane winds. In the Dinaric region, sanitary felling was caused by freezing rain and biotic factors. Croatian uneven-aged conifer forests are generally characterized by higher levels of biodiversity in terms of tree age and species composition and thus suffer lower damage, as was again proven by the part of this research. The monitoring of bark beetles is a mandatory necessity, which can, again, give us valuable data in the upcoming years. Climatic extremes recorded in the last decade in lowland and Dinaric Croatia included dry years, wet years, pest gradations, freezing rains, and events with hurricane wind speeds. The damage caused to forest ecosystems was sudden tree death, windbreaks, windthrows, fractures, and deformations of crowns and entire trees. Sustainable forest management is deeply rooted in the principles of Croatian forestry. However, disturbances that cause a decrease in tree vitality and mortality over large areas represent a serious challenge for sustainable forest management planning. Sanitary felling is carried out in forest stands affected by natural disturbances caused by winds, droughts, fires, floods, pollution, and insects, or the adverse effects of abiotic and/or biotic factors. The ratio of the volume of sanitary felling to the total annual felling of trees shows us the size and intensity of the disturbance in the ecosystem. In order to rehabilitate the area affected by disturbances, the amounts of sanitary felling can be many times higher than the planned amounts of felling. The damage affects certain specific forest ecosystems, which disrupts biodiversity at the ecosystem level. Even though resilience is considered a key concept mandatory for future forestry, it is still evident that more than one challenge is before us, not just in the afforestation of damaged sites that maintain biodiversity, but also in the general lack of the workforce.