Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2022 May 26.
Published in final edited form as: Biol Conserv. 2018 Jul 31;226:81–91. doi: 10.1016/j.biocon.2018.07.014

Forest structure following natural disturbances and early succession provides habitat for two avian flagship species, capercaillie (Tetrao urogallus) and hazel grouse (Tetrastes bonasia)

Mareike Kortmann a,*, Marco Heurich b,c, Hooman Latifi d,e, Sascha Rösner f, Rupert Seidl g, Jörg Müller a,b, Simon Thorn a
PMCID: PMC7612776  EMSID: EMS145170  PMID: 35633892

Abstract

Boreal and mountainous forests are a primary focus of conservation efforts and are naturally prone to large-scale disturbances, such as outbreaks of bark beetles. Affected stands are characterised by biological legacies which persist through the disturbance and subsequent succession. The lack of long-term monitoring data on post-disturbance forest structure precludes understanding of the complex pathways by which natural disturbances affect forest structure and subsequently species presence. We analysed the response of capercaillie (Tetrao urogallus) and hazel grouse (Tetrastes bonasia) to bark beetle infestations. We combined high-resolution airborne light detection and ranging (LiDAR) with a 23-year time series of aerial photography to quantify present-day forest structure and stand disturbance history. Species presence was assessed by collecting droppings of hazel grouse and capercaillie in a citizen science project. Structural equation models showed that the probability of hazel grouse presence increased with increasing disturbance, and the probability of both hazel grouse and capercaillie presence increased with succession. Indirect effects of bark beetle infestations, such as a reduced abundance of deciduous trees and an enhanced herb layer cover, were positively associated with capercaillie presence. Decreasing canopy cover increased the probability of hazel grouse presence. The high temporal and spatial heterogeneity of bark beetle infestations created forest structures that meet the contrasting habitat requirements of both, capercaillie and hazel grouse. This heterogeneity resulted from biological legacies such as decomposing snags, and the simultaneous regrowth of natural regeneration. A benign-neglect strategy towards bark beetle infestations could hence foster capercaillie and hazel grouse in mountainous forests.

Keywords: Citizen science, Forest succession, Habitat selection, LiDAR, Natural disturbance, Tetraoninae

1. Introduction

Mountainous forest ecosystems provide numerous ecosystem services, such as merchantable timber, climate regulation, water purification and recreation (Kulakowski et al., 2016; Leverkus et al., 2015). Because of intensive human use, European mountainous forests underwent multiple alterations over the last centuries, which resulted in a loss of old-growth forest structures and biodiversity (Bengtsson et al., 2000; Paillet et al., 2010). Forest managers have also aimed to minimize the damage created by natural disturbances, such as windstorms, wildfire and insect outbreaks (Schelhaas et al., 2003). Nonetheless, climate change has led to an increase in the frequency and intensity of natural disturbances during the last decades (Schelhaas et al., 2003; Seidl et al., 2014). The economically most important biotic disturbance agent in Central Europe is the European spruce bark beetle (Ips typographus L.), which infests stands of Norway spruce (Picea abies (L.) Kast) (Wermelinger, 2004). Outbreaks of I. typographus have severely affected commercial forests and have led to significant economic losses (an average of 2.9 million m3 timber damaged per year between 1950 and 2001 and 14.5 million m3 between 2001 and 2010 in Europe) (Schelhaas et al., 2003). Moreover, I. typographus is expanding northwards and into higher elevations because of a warming climate, making mountainous forests more susceptible to bark beetle outbreaks (Jönsson et al., 2009; Seidl and Rammer, 2017).

Bark beetle infestations can have multiple positive effects on forest biodiversity, ranging from promoting abundances of individual species (e.g. Clark et al., 2013; Kortmann et al., 2017) to creating habitat for entire communities (Beudert et al., 2015; Lehnert et al., 2013). Such positive effects are predominantly mediated by increases in the amount of dead wood, solar insolation, the removal of highly competitive, dominant trees and enhanced structural heterogeneity (Swanson et al., 2011). However, impacts on tree species composition and forest structure following disturbances are extremely variable (Swanson et al., 2011). The response of species to disturbances is driven by multiple factors, which can affect changes in species abundance in several ways, for example by behavioural adaptations, increased recruitment, enhanced mortality rates (Pausas and Parr, 2018; Banks et al., 2012) or post-disturbance dispersal (Pierson et al., 2013). Also, specific structural attributes of succession of naturally disturbed forests and the associated resources e.g. food availability, can have a strong influence on the presence of species (Swanson et al., 2011). For example, insectivorous birds, that are specialized on bark-dwelling invertebrates, generally benefit from bark beetle infestations, while other avian foraging guilds may remain unaffected (Drever et al., 2009). The population densities of six species of woodpeckers in British Columbia increased rapidly during outbreaks of the mountain pine beetle (Dendroctonus ponderosae Hopkins) because of the increased amount of dead wood and the rapidly increasing abundance of arthropods (Edworthy et al., 2011). However, the effects of bark beetle infestations on species occurrence are complex, and an understanding of post-disturbance habitat suitability for many flagship species of conservation in Europe is still lacking.

Capercaillie (Tetrao urogallus L.) and hazel grouse (Tetrastes bonasia L.) inhabit boreal and mountainous forests in Europe (Storch, 1993). The capercaillie is a flagship species for forest conservation in Europe, and is listed in Annex I, II and III of the European Council Directive (Birds Directive) on the protection of wild birds (79/409/EEC), protecting birds that are particularly threatened and postulating the designation of “Special Areas for Conservation”. Annex II and III restrict hunting to certain conditions and designated hunting periods. The hazel grouse is listed in the Annex I of the Birds Directive. Both tetraonids are mainly threatened by habitat loss, forest fragmentation, human disturbance and climate change (Åberg et al., 1995; Braunisch et al., 2014; Rösner et al., 2014a, 2014b; Sahlsten et al., 2010; Selås et al., 2011; Storch, 2007a). Previous studies investigating the habitat preferences of capercaillie and hazel grouse report contrasting preferences for successional stages and the associated forest structures by the two species (Table 1).

Table 1. Forest structures that are associated with capercaillie (Tetrao urogallus L.) and hazel grouse (Tetrastes bonasia L.) in European forests.

Forest structure Capercaillie (Tetrao urogallus) Hazel grouse (Tetrastes bonasia)
Quality Reference Quality Reference
Canopy cover Sparse Sachot et al. (2003) High Vauhkonen and Imponen (2016), Ludwig and Klaus (2016)
Intermediate (30–60%) Bollmann et al. (2005), Graf et al. (2009) Diverse Schäublin and Bollmann (2011)
Moderate (50–60%) Storch (2002)
Incomplete Suter et al. (2002)
Vertical forest heterogeneity Two or more vertical strata Vauhkonen and Imponen (2016) Multiple vertical strata Vauhkonen and Imponen (2016)
Multi-story tree layer Suter et al. (2002) Well structured Müller et al. (2009)
High structural heterogeneity González et al. (2012) Multi-layered Pfandl et al. (2013)
Herb layer Low Sachot et al. (2003) High Vauhkonen and Imponen (2016)
Average coverage Vauhkonen and Imponen (2016) Close to 50% coverage Sachot et al. (2003)
Rich in ericaceous shrubs Bañuelos et al. (2008) Rich field layer Åberg et al. (2003)
Shrub layer Mixed regeneration Teuscher et al. (2011) Dense Schäublin and Bollmann (2011)
<25% coverage Storch (2002) Dense Ludwig and Klaus (2016)
Low Mikolás et al. (2017)
Deciduous tree species Low amounts Thiel et al. (2007) Present (species presence independent of increase) Schäublin and Bollmann (2011)
Open in a new tab

The natural habitat of both species is periodically affected by natural disturbances, such as bark beetle outbreaks, which significantly alter forest structure. Specifically, hazel grouse occurs predominantly in early successional stages in which many deciduous trees and shrubs serve as important food resources and shelter (Zellweger et al., 2013). In contrast, capercaillie occurs mostly in coniferous and mixed late seral forests that are characterised by a herb layer dominated by ericaceous plants and open areas that are utilized for courtship behaviour (Bañuelos et al., 2008). A lack of detailed information on disturbance-mediated changes in forest structures often hampers the quantitative analysis of the effects of disturbances and succession on capercaillie and hazel grouse (Mikolás et al., 2017). As the result of bark beetle infestations, mature spruce trees die, resulting in an opening of the canopy and in enhanced growth of regeneration and shrub layers (Swanson et al., 2011). Capercaillie, which prefers open forest structures, might profit from such canopy opening. In contrast, hazel grouse might benefit from increased understorey growth in the years following a disturbance.

In this study, we combined temporal and spatial information from high-resolution aerial photography, airborne light detection and ranging (LiDAR), and presence data of both capercaillie and hazel grouse. We used structural equation modelling to test the effects of bark beetle infestations on the probability of capercaillie and hazel grouse presence mediated by changes in canopy cover, vertical forest heterogeneity, shrub and herb layer cover, and abundance of deciduous trees. We hypothesized that bark beetle infestations have a major influence on capercaillie and hazel grouse presence, e.g. a positive effect of increased canopy opening on capercaillie presence and a positive effect of enhanced lower vegetation layers on hazel grouse presence. We further hypothesized that effects are primarily mediated by changes in forest structure rather than by direct effects of disturbance.

2. Material and methods

2.1. Study area

All data were collected in the Bavarian Forest National Park in south-eastern Germany (48°54′N, 13°29′E). The park is part of one of the largest contiguous forested regions in Central Europe, the Bohemian Forest Ecosystem, which also comprises the Šumava National Park in the Czech Republic. The national park consists of mixed, mountainous forests in areas below 1150 m a.s.l., composed mainly of Norway spruce, European beech (Fagus sylvatica L.) and silver fir (Abies alba Mill.). High montane forests in areas above 1150 m a.s.l. are naturally dominated by Norway spruce (Bässler et al., 2010). Annual precipitation ranges between 1200 mm and 1800 mm, and mean annual temperature ranges between 3.8 °C and 5.8 °C.

During the 18th and 19th centuries, the entire area was disturbed several times by windstorms and subsequent bark beetle infestations (Brůna et al., 2013). Since the foundation of the Bavarian Forest National Park in 1970, a benign-neglect strategy towards natural disturbances has allowed the bark beetle to kill mature spruces without human intervention. As a result, the study area exhibits a diverse set of conditions ideally suited to study the effect of disturbances on capercaillie and hazel grouse, with areas disturbed and undisturbed by bark beetles distributed across the entire national park (Beudert et al., 2015; Thorn et al., 2016).

2.2. Species data

The capercaillie population in the Bavarian Forest National Park is, besides that of the Black Forest, one of the last populations in European low mountain ranges (Rösner et al., 2014a; Storch, 2007b). It declined severely during the 20th century, with only approximately 16 birds left in 1985 (Rösner et al., 2014b). In the 1980s, an intense breeding program released > 1300 individuals. Based on population genetic measures, Rösner et al. (2014a) estimated the current population size at approximately 500 individuals for the entire Bohemian Forest Ecosystem. It is thought that the hazel grouse population in the Šumava National Park, i.e. the adjacent Czech part of the Bohemian Forest Ecosystem, is stable or has even been slightly increasing since the middle of the 20th century (Klaus, 2007). Today, the Bohemian Forest Ecosystem contains a large portion of the hazel grouse population remaining outside of the Alps (Müller et al., 2009).

Capercaillie and hazel grouse droppings were collected year-round between 2009 and 2011 by a group of approximately 70 trained volunteers in a citizen science project (Rösner et al., 2014a). The volunteers systematically searched for capercaillie droppings twice per season (two times in summer (May–September) and two times in winter (October–April)) in areas delineated by grids (for details, see Rösner et al., 2014a, 2014b). Furthermore, we included all observations of birds (visual and audible) in the Bavarian Forest National Park from 2010 and 2011, which are regularly gathered by rangers and other staff. In total, we obtained 433 records of capercaillie and 139 of hazel grouse (Fig. 1).

Fig. 1.

Fig. 1

Open in a new tab

Presence records of capercaillie (Tetrao urogallus L.), and hazel grouse (Tetrastes bonasia L.) based on chance observations and droppings collected between 2009 and 2011 in the Bavarian Forest National Park, Germany (black contour line represents the park border). The map in the upper right corner shows the location of the study area in Germany at the border to the Czech Republic. The interval between elevation contour lines is 100 m, and spans from 800 m to 1400 m a.s.l.

2.3. Forest structures

We used LiDAR data to acquire information about forest structure in a 60-m radius around every presence point, and added 987 additional random points, evenly distributed over the entire study area as pseudo-absence data for comparison. Because the study area is highly heterogeneous, we selected a radius of 60 m to obtain small-scale habitat parameters on the stand level. The study area was scanned with an airborne Riegl LMS-Q680i sensor in June 2012 from a height of 650 m with an average density of 30 points m−2 (Latifi et al., 2015). The raw data including the collected waveforms were decomposed in XYZ coordinates of individual returns by Gaussian functions (Reitberger et al., 2009). Point information was aggregated into 10 m × 10 m grids. For every 10 m × 10 m grid cell LiDAR parameters were calculated in vegetation layers differentiated by height (herb layer, shrub layer, third tree layer, second tree layer and first tree layer).

We selected canopy cover [%], vertical forest heterogeneity [m], herb layer cover [%], shrub layer cover [%], and number of deciduous trees [%] as forest structures of relevance for capercaillie and hazel grouse presence, following existing literature (Table 1). Canopy cover, herb layer cover und shrub layer cover are predictive extrapolations of model-based calculations. They were calculated based on LiDAR metrics using the original high-density point cloud (for details, see Latifi et al., 2017). Preselection of the LiDAR metrics was based on the definition of the metrics and the previous experiences on their association with the respective layer across the similar study site (for details see Latifi et al., 2015). Canopy cover was defined as vegetation density in the upper two-thirds of the stand top height (originally suggested by Næsset (2004) and further verified by Latifi et al. (2015)).

The herb layer was defined as vegetation < 1.5 m in height, including grasses, dwarf shrubs and small saplings. According to Latifi et al. (2017), the calculation of the herb layer cover included the percentage of LiDAR returns located between 0 m and 1.5 m above ground, the height skewness of the returns located between 0 m and 1.5 m, a cover understorey metric n(height<2m)n(height<60m), and the interaction between habitat types and the cover understorey metric. Shrub layer was defined as vegetation layer between 1.5 m and 5 m in height, including shrubs and large saplings. The calculations of the shrub layer cover included the height skewness of the returns located between 1.5 m and 5 m and a n(height<1.5m)n(height<5m), Vertical forest heterogeneity was represented by the standard deviation of the mean height of all LiDAR returns in the 10 m × 10 m grids falling within the 60-m radius around each presence point. A high standard deviation was assumed to indicate a high structural heterogeneity, i.e. a large variety of tree heights within the 60-m radius, while a small standard deviation indicates that trees are uniformly sized (for details, see Latifi et al., 2015). We used normalized cut segmentation to extract single trees and classified tree species based on properties of the LiDAR point cloud, for which the detailed procedure is described by Yao et al. (2012, 2014). The percentage of deciduous trees was calculated based on the crown volume of all single trees in the 60-m surrounding of each presence and absence record. The obtained information on forest structures was subsequently modelled as a function of disturbance and natural succession.

2.4. Bark beetle infestations

Bark beetle infestations were documented for every year since 1988 in summer or early autumn by means of colour infrared aerial photographs, with a ground resolution of 20 cm (Heurich et al., 2010; Kautz et al., 2011). CIR photographs from 1988 to 2011 were visually interpreted to obtain information on bark beetle infestations. Based on the acquired information a grid system was established. Every 30 m × 30 m grid was classified according to whether it was disturbed by bark beetles or not (Kautz et al., 2011). Grid cells that had never been disturbed by bark beetle infestations were classified as undisturbed forest. Beetle-affected grid cells were classified as disturbed forest, and the year of bark beetle disturbance was extracted. Time since the disturbance is hereafter referred to as succession. We restricted our analysis to grids that were classified as undisturbed forest and grids that were disturbed by bark beetles and excluded all other kinds of disturbances, i.e. salvage logging and windthrow. The dataset included only 13 grids that were identified as salvage logged or disturbed by windthrow. Due to the small extent we removed them from the final dataset. This data set of forest disturbance provides comprehensive insight into the forest dynamics of the Bavarian Forest National Park over the last 23 years (Seidl et al., 2016).

2.5. Statistical analyses

All statistical analyses were conducted in R 3.3.1. (R Core Team, 2013). We calculated the percentage of the different stand types (disturbed forest, undisturbed forest) for all presence records and the randomly selected plots in their 60-m surrounding.

To test the influence of bark beetle infestations and succession on forest structures as well as their influence on the probability of capercaillie and hazel grouse presence, we used piecewise structural equation modelling (SEM) from the piecewiseSEM package (Lefcheck, 2016). The SEM included different individual models for every response variable to model the single relationships between two variables (Shipley, 2009, 2013). To test the interrelationships amongst different processes, i.e. the paths between all included variables, the individual models were combined in one SEM. Hence, the SEM tested the influence of bark beetle infestations on forest structures and the effect of forest structures on capercaillie and hazel grouse presence. First, we used generalized linear models (GLMs) from the lme4 package (Bates et al., 2015) to test the influence of forest structures, disturbance and succession on capercaillie and hazel grouse presence/absence. Elevation above sea level was included in the GLMs as a covariate to control for varying environmental conditions. We used a second set of linear models (LMs) from the lme4 package to test the influence of the number of disturbed forest stands, succession and elevation on forest structures (canopy cover [%], vertical forest heterogeneity [m], herb layer cover [%], shrub layer cover [%] and deciduous trees [%]) (Table 2).

Table 2.

Statistical models that were combined in piecewise structural equation modelling (SEM). We used generalized linear models (GLMs) to model the effect of forest structures on the presence (0/1) of capercaillie (Tetrao urogallus L.) and hazel grouse (Tetrastes bonasia L.). Linear models (LMs) were used to test the influence of the amount of disturbed forest stands, succession and elevation on the different forest structures.

Model type Response variable Predictors
GLM Capercaillie presence/absence Canopy cover [%] + vertical forest heterogeneity [m] + herb layer cover [%] + shrub layer cover [%] + deciduous trees [%] + disturbed forest stands [m2] + succession [years] + elevation [m a.s.l.]
GLM Hazel grouse presence/absence Canopy cover [%] + vertical forest heterogeneity [m] + herb layer cover [%] + shrub layer cover [%] + deciduous trees [%] + disturbed forest stands [m2] + succession [years] + elevation [m a.s.l.]
LM Canopy cover [%] Disturbed forest stands [m2] + succession [years] + elevation [m a.s.l.]
LM Vertical heterogeneity [m] Disturbed forest stands [m2] + succession [years] + elevation [m a.s.l.]
LM Herb layer cover [%] Disturbed forest stands [m2] + succession [years] + elevation [m a.s.l.
LM Shrub layer cover [%] Disturbed forest stands [m2] + succession [years] + elevation [m a.s.l.]
LM Deciduous trees [%] Disturbed forest stands [m2] + succession [years] + elevation [m a.s.l
Open in a new tab

To consider possible non-linear effects of forest structures on capercaillie and hazel grouse presence, we additionally calculated generalized additive models (GAMs). The GAMs included the same variables as the GLMs used in the SEM. To consider possible non-linear effects of bark beetle outbreaks and succession on forest structures, we included variables with expected non-linear (i.e. unimodal) effects as second-degree polynomials in the LMs used in the SEM via the r-function “poly”. Non-linear effects were included in the final model, after carefully evaluating model outputs for overfitting.

To test for possible correlations between the variables, we calculated the variance inflation factor of all variables with the “vifstep” function from the usdm package with a threshold of 5 (Naimi, 2015). The maximum variance inflation factor was 3.30 for canopy cover (Table A.2), hence, we included all variables in the analyses. Furthermore, we estimated the degree of correlation of all model residuals with geographic coordinates by means of spline correlograms (Bjørnstad and Falck, 2001) to detect possible spatial autocorrelations. We did not detect any spatial dependency in our model outputs (Fig. A.1).

3. Results

Bark beetle infestations occurred regularly during the last 23 years in our study area, with peaks 5 and 10–15 years ago (Fig. 2). Capercaillie and hazel grouse presence was highest in forest stands 10–15 years after disturbance (Fig. 2; see Fig. 5e and 6f for model results, controlled for elevation). Results of the SEMs showed that previously disturbed forests had a positive effect on the probability of hazel grouse presence (p < 0.001; z = 3.86). Additionally, succession (time since disturbance) was positively associated with an increasing probability of both capercaillie (p < 0.001; z = 4.52) and hazel grouse (p = 0.016; z = 2.39) presence (Fig. 3). Furthermore, the probability of capercaillie presence increased with increasing elevation (p < 0.001; z = 8.82).

Fig. 2.

Fig. 2

Open in a new tab

Disturbance history of A) plots with capercaillie (Tetrao urogallus L.) presence, B) plots with hazel grouse (Tetrastes bonasia L.) presence and C) randomly selected control plots. Bars represent the number of plots disturbed in the respective year calculated as weighted mean in the 60-m radius of each plot. Year 1, 2 and 3 since disturbance refer to year 2009 until 2011, where species data were gathered.

Fig. 5.

Fig. 5

Open in a new tab

Partial effects of a) vertical heterogeneity, b) canopy cover, c) elevation, d) disturbed forest stands and e) succession, isolated from generalized additive models on the presence of hazel grouse (Tetrastes bonasia). Y-axis presents the isolated effects of the respective predictor on the presence of hazel grouse. Non-significant variables are not displayed (please see Appendix Table A.1 for full model results).

Fig. 6.

Fig. 6

Open in a new tab

Partial effects of a) vertical heterogeneity, b) canopy cover, c) shrub layer, d) herb layer, e) elevation, f) succession and g) deciduous trees, isolated from generalized additive models on the presence on the presence of capercaillie (Tetrao urogallus). Y-axis presents the isolated effects of the according predictor on the presence of capercaillie. Non-significant variables are not displayed (please see Appendix Table A.1 for full model results).

Fig. 3.

Fig. 3

Open in a new tab

Structural equation model with the amount of disturbed areas and succession (time since disturbance) on the left, forest structures in the centre, and presence of capercaillie (Tetrao urogallus L.) and hazel grouse (Tetrastes bonasia L.) on the right. Black arrows illustrate a positive association with a respective response variable; grey arrows illustrate a negative association with a respective response variable. Non-linear effects of succession on herb layer cover, vertical heterogeneity and shrub layer cover are shown with dotted arrows. The predictions of these effects are shown in Fig. 4. Arrows are labelled with the corresponding z-values and significance levels (***, p ≤ 0.001; **, 0.001 < p ≤ 0.01; *, 0.01 < p ≤ 0.05). Z-values of second-degree polynomials of non-linear effects are italicised. Arrow width corresponds to z-values (see Table A.1 for coefficients).

An increased share of disturbed forests (p < 0.001; z = −3.48) and succession (p < 0.001; z = 11.39) had a significantly negative effect on canopy cover, which in turn had a significant positive influence on hazel grouse presence (p = 0.003; z = −3.02). Conversely, herb layer cover increased with increasing disturbance (p = 0.009; z = 2.61), but decreased with progressing succession (Fig. 4). Herb layer cover was positively associated with capercaillie presence (p < 0.001; z = 6.05). The number of deciduous trees decreased with increasing disturbance (p < 0.001; z = −4.06), which in turn positively affected the presence of hazel grouse (p < 0.001; z = 3.75), but negatively affected the presence of capercaillie (p < 0.001; z = −6.45). Shrub layer cover significantly increased with succession (Fig. 4), ultimately resulting in a decreasing probability of capercaillie presence (p < 0.001; z = −4.75). Succession had a direct positive effect on the presence of capercaillie (p < 0.001; z = 4.52) (Fig. 3).

Fig. 4.

Fig. 4

Open in a new tab

Non-linear effects of succession on a) herb layer cover, b) vertical heterogeneity and c) shrub layer cover based on second-order polynomials in a structural equation model.

4. Discussion

Natural disturbance by bark beetle infestations and the following succession created heterogenous forest structures that meet the contrasting habitat requirements of both capercaillie and hazel grouse. Hazel grouse presence was positively affected by the disturbance events that occurred in our study area during the last 23 years. Succession, i.e. time since disturbance, led to an increase in both capercaillie and hazel grouse presence.

4.1. Impacts of bark beetle infestations and succession on forest structures

Succession influenced all analysed indicators of forest structure. Conversely, only canopy cover (negative), herb layer cover (positive) and deciduous trees (negative) were directly influenced by disturbance. Trees affected by bark beetles lose their needles within the first years after an outbreak (Mikkelson et al., 2016), and snags typically collapse after 5–25 years (Mitchell and Preisler, 1998), which results in a gradual decrease in canopy cover over several years. Hence, besides the disturbance itself, also post-disturbance decomposition processes and succession have impacts on forest structure. Specifically, our results showed that canopy cover decreased with increasing bark beetle activity, and decreased during the first years after disturbance. Furthermore, natural regeneration takes several decades to establish a closed canopy (Edburg et al., 2012), depending on site factors and the availability of seed sources (Swanson et al., 2011). During this stage, the increased light availability and enhanced temperature amplitude facilitate the growth of herbaceous plants (Fontaine et al., 2010). After a few years, these shade-intolerant herbaceous species are surpassed by natural regeneration of spruces and other perennial species (Swanson et al., 2011). We found that herb layer cover is higher on disturbed sites due to overstorey loss, but it decreases over the course of succession. In contrast, shrub layer cover increased with ongoing succession (Fig. 3).

The number of deciduous trees was lower in disturbed forests and decreased over the course of succession. Lower numbers of deciduous trees in disturbed forest stands could be explained by the predominant presence of bark beetle infestations in stands with higher numbers of spruce. On the other hand, the number of deciduous trees decreases during succession because spruce saplings dominate the regeneration (Zeppenfeld et al., 2015). Our results are in line with those of Fischer et al. (2015), who showed that the abundance of light-demanding deciduous species increased only slightly after a stand-replacing disturbance in Norway spruce mountain forests. In contrast to wildfires, bark beetle infestations typically leave understorey vegetation largely intact. Consequently, spruce saplings are already present for recruitment when mature spruces in the overstorey die because of bark beetle infestations (Swanson et al., 2011). Under strong reduction in canopy cover, spruce saplings are more competitive than beech saplings. Nevertheless, it should be considered that Zeppenfeld et al. (2015) focussed on forest stands at higher elevations that were already dominated by Norway spruce before the disturbance, whereas our study area consists also of mixed forest stands with European beech, silver fir and Norway spruce. In these mixed forest stands the successional change in tree species composition differs from spruce dominated forests. Small numbers of infested Norway spruce canopy trees result in only subtle canopy opening after a bark beetle infestation and hence little advantage for spruce saplings. The fact that the most intense bark beetle infestations in our study area occurred in higher elevation area, characterised by mountain spruce forests, might explain the overall decrease in deciduous trees after a bark beetle infestation found here.

Our results showed that the LiDAR-derived stand vertical heterogeneity decreased during succession. This finding is in contrast with the framework of Donato et al. (2012), who predicted that stand-replacing natural disturbances create vertically heterogeneous forest stands with continuously high levels of complexity across all successional stages. Early seral stands disturbed by bark beetles are composed of clusters of regenerating trees and snags of beetle-killed spruces (Wild et al., 2014). This initial heterogeneous pattern homogenized over the 23-year course of succession studied here, as snags collapsed and natural regeneration increased in density. While patches of bark beetle disturbance are very heterogeneous at the landscape scale (Senf and Seidl, 2017), our results indicate a loss of heterogeneity at smaller spatial scales. Scale is thus of crucial importance for quantifying the effects of disturbances on forest structure.

4.2. Impacts of bark beetle infestations on capercaillie and hazel grouse

Changes in forest structure had major impacts on both capercaillie and hazel grouse. According to our results, a decreased canopy cover has a positive effect on the probability of hazel grouse presence, which is in line with previous studies (Table 1) describing open to semi-open forest stands with dense understorey as preferred habitat for hazel grouse. The increased canopy opening and concomitant increase in solar radiation reaching the forest floor in disturbed forests (Fontaine et al., 2010) can promote the establishment of early-seral species (Koivula and Spence, 2006; Rost et al., 2012; Thom et al., 2017). Early-seral tree species, such as rowan (Sorbus aucuparia L.) and willow (Salix sp.), serve as important food resources for hazel grouse (Schäublin and Bollmann, 2011). Hence, the presence of such tree species in the shrub layer increases the habitat quality for hazel grouse (Zellweger et al., 2014). For instance, Schäublin and Bollmann (2011) conclude that a minimum amount of ~1 tall rowan per ha had a strong positive effect on hazel grouse presence in wintering habitats in the Alps. Such important changes in tree and shrub species composition might also explain the positive direct impacts of the number of beetle-killed grid cells that could not be explained by forest structure (uppermost path in Fig. 3). Although LiDAR data allows a differentiation between deciduous and coniferous trees, specific tree species like willow and rowan, are not represented in our forest structural data and are therefore expressed by the direct impacts of disturbance on capercaillie and hazel grouse presence. Our results demonstrate that the number of deciduous trees and canopy cover are the main factors that determine the probability of hazel grouse presence in our study area. Our assumption that shrub layer cover should affect habitat quality for hazel grouse could not be verified. Zellweger et al. (2014) confirmed that horizontal heterogeneity explains more of the variance in hazel grouse occurrence than vertical heterogeneity. However, the reference-area in their analysis was approximately 100 times larger than in our study. At small scales such as in our study, the horizontal heterogeneity may have another and maybe different effect. Considering that after bark beetle infestations there are always remaining unaffected spruce stands on a landscape scale, the canopy cover on a local scale might describe similar forest structures as the horizontal heterogeneity in a larger reference area. Furthermore, canopy cover affects the species composition in the shrub layer and enhances the abundance of early-seral tree species as described above. Hence, the species composition of the shrub layer appears to be of higher importance for habitat suitability than the cover of the shrub layer and we cannot detect species composition by means of LiDAR, which might explain the missing link between shrub layer cover and the probability of hazel grouse presence in our study.

The probability of capercaillie presence decreased with increasing number of deciduous trees (Fig. 2). This is in line with the findings of Thiel et al. (2007), who described the importance of conifer needles for capercaillie during winter. Conifer needles have low energy content compared to deciduous leaves and therefore need to be highly abundant to serve as a suitable food resource. Furthermore, in the Bavarian Forest National Park, forest stands with a high proportion of deciduous trees are mixed, mountainous forests, and are often characterised by a dense shrub layer of European beech regeneration (Bässler et al., 2010). Such stands had a lower probability of capercaillie presence, particularly because increasing shrub layer cover was negatively associated with capercaillie presence (Fig. 3). According to the results of our modelling, suitable habitats for capercaillie are stands with herb layer cover > 35%, shrub layer cover < 15% and a canopy cover < 60% (Fig. 6). This is in line with results of other studies (Table 1) that found a preference of the species for sparse to intermediate canopy cover and a well-developed herb layer.

The probability of capercaillie presence increased over the course of post-disturbance succession. Decay and breakage of standing and downed dead wood lead to more open sites interspersed with groups of natural spruce regeneration, which represents increasingly suitable habitat for capercaillie. Especially buds and new shoots of young spruce trees serve as important winter forage for capercaillie (Braunisch and Suchant, 2013). During summer, an increased herb layer cover provides suitable feeding resources owing to the growth of dwarf shrubs. This is in line with the results of several other studies that underline the importance of dwarf shrubs such as bilberry (Vaccinium myrtillus L.) for capercaillie (Sachot et al., 2003; Storch, 1993). In contrast, the probability of capercaillie presence decreased in stands with a dense shrub layer. Indeed, Storch (2002) described that natural regeneration covering > 75% makes habitats unsuitable for capercaillie. A possible reason for this finding is given by Sachot et al. (2003), who showed that capercaillie avoided stands with a dense understorey, whereas a sparse understorey allows them to rapidly take flight if necessary. During succession, herb layer cover decreases, whereas shrub layer cover increases. Consequently, and according to Sachot et al. (2003), with ongoing succession, the forest structure of disturbed forest stands should become unsuitable for capercaillie. However, our results do not support this conclusion. Nevertheless, this finding might change if successional trajectories beyond 23 years are investigated (i.e. the establishment of mature forests with closed canopies). In our study area and over the timespan of 23 years, other factors, such as food resources, are the primary factors affecting capercaillie presence. Considering that capercaillie occurs mainly at higher elevations in our study area (Teuscher et al. 2011), where spruce regeneration is patchy, the positive effect of spruce needles and shoots as food resources is more decisive than the negative effect of an increasing shrub layer. These differences between the forest types in our study area might also explain the negative impact of deciduous trees on the presence of capercaillie. As described above, stands with a high number of deciduous trees are primarily mixed, mountainous forests in the Bavarian Forest National Park and are characterised by a dense understorey. High montane forests, on the other hand, are dominated by spruce with sparse understorey and well-developed herb layer, hence representing highly suitable habitat for capercaillie. This is confirmed by our results that capercaillie presence increased with elevation (Table A.1).

5. Conclusion

Our results confirm that bark beetle outbreaks enhance the habitat suitability for both capercaillie and hazel grouse in mountainous forests. Whereas traditional management recommendations are based on forest management strategies that actively modify forest structure, our study suggests that bark beetle infestation and succession can be used as a passive alternative for creating habitat for capercaillie and hazel grouse by altering the forest structure of previously managed forest stands.

Forest structures are affected in contrasting ways by natural disturbance and succession. In turn, the presence of the two species was differentially related to forest structure (Fig. 2). Our findings highlight that a key impact of natural disturbances is their heterogeneous effect on forest canopies, creating a complex mosaic of structures in space and time. Sachot et al. (2003) previously claimed that for the coexistence of capercaillie and hazel grouse, a mosaic of different habitat types is necessary. Here we show that the irregular occurrence of bark beetle infestations in combination with different trajectories of recovery in early seral stands creates high spatial and temporal heterogeneity (Swanson et al., 2011) and meets the habitat demands of both capercaillie and hazel grouse: dense, shrubby forest stands for hazel grouse as well as open areas with dwarf shrubs and grasses, interspersed with young conifers for capercaillie. Future studies should target the long-term effects of natural disturbances on forest structures and biodiversity.

Supplementary Material

Appendix

Acknowledgements

MK received support from the scholarship program of the German Federal Environmental Foundation (20017/474) and RS from a START grant of the Austrian Science Fund FWF (Y 895-B25). We thank Karen A. Brune for linguistic revision of the manuscript and three anonymous reviewers who helped to improve the manuscript with their valuable comments.

References

  1. Åberg AJ, Jansson G, Swenson JE, Angelstam P. The effect of matrix on the occurrence of hazel grouse (Bonasa bonasia) in isolated habitat fragments. Ecology. 1995;103:265–269. doi: 10.1007/BF00328613. [DOI] [PubMed] [Google Scholar]
  2. Åberg J, Swenson JE, Angelstam P. The habitat requirements of hazel grouse (Bonasa bonasia) in managed boreal forest and applicability of forest stand descriptions as a tool to identify suitable patches. For Ecol Manag. 2003;175:437–444. [Google Scholar]
  3. Banks SC, Blyton MDJ, Blair D, McBurney L, Lindenmayer DB. Adaptive responses and disruptive effects: how major wildfire influences kinship-based social interactions in a forest marsupial. Mol Ecol. 2012;21:673–684. doi: 10.1111/j.1365-294X.2011.05282.x. [DOI] [PubMed] [Google Scholar]
  4. Bañuelos M-J, Quevedo M, Obeso J-R. Habitat partitioning in endangered Cantabrian capercaillie Tetrao urogallus cantabricus. J Ornithol. 2008;149:245–252. [Google Scholar]
  5. Bässler C, Müller J, Dziock F. Detection of climate-sensitive zones and identification of climate change indicators: a case study from the Bavarian Forest National Park. Folia Geobot. 2010;45:163–182. [Google Scholar]
  6. Bates D, Mächler M, Bolker BM, Walker SC. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67:1–48. [Google Scholar]
  7. Bengtsson J, Nilsson SG, Franc A, Menozzi P. Biodiversity, disturbances, ecosystem function and management of European forests. For Ecol Manag. 2000;132:39–50. [Google Scholar]
  8. Beudert B, Bässler C, Thorn S, Noss R, Schröder B, Dieffenbach-Fries H, Foullois N, Müller J. Bark beetles increase biodiversity while maintaining drinking water quality. Conserv Lett. 2015;8:272–281. [Google Scholar]
  9. Bjørnstad ON, Falck W. Nonparametric spatial covariance functions: estimation and testing. Environ Ecol Stat. 2001;8:53–70. [Google Scholar]
  10. Bollmann K, Weibel P, Graf RF. An analysis of central Alpine capercaillie spring habitat at the forest stand scale. For Ecol Manag. 2005;215:307–318. [Google Scholar]
  11. Braunisch V, Suchant R. Aktionsplan Auerhuhn Tetrao urogallus im Schwarzwald: Ein integratives Konzept zum Erhalt einer überlebensfahigen Population. Vogelwelt. 2013;41:29–41. [Google Scholar]
  12. Braunisch V, Coppes J, Arlettaz R, Suchant R, Zellweger F, Bollmann K. Temperate mountain forest biodiversity under climate change: compensating negative effects by increasing structural complexity. PLoS ONE. 2014;9:e97718. doi: 10.1371/journal.pone.0097718. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Brůna J, Wild J, Svoboda M, Heurich M, Müllerová J. Impacts and underlying factors of landscape-scale, historical disturbance of mountain forest identified using archival documents. For Ecol Manag. 2013;305:294–306. [Google Scholar]
  14. Clark DA, Anthony RG, Andrews LS. Relationship between wildfire, salvage logging, and occupancy of nesting territories by northern spotted owls. J Wildl Manag. 2013;77:672–688. [Google Scholar]
  15. Donato DC, Campbell JL, Franklin JF. Multiple successional pathways and precocity in forest development: can some forests be born complex? J Veg Sci. 2012;23:576–584. [Google Scholar]
  16. Drever MC, Goheen JR, Martin K. Species–energy theory, pulsed resources, and regulation of avian richness during a mountain pine beetle outbreak. Ecology. 2009;90:1095–1105. doi: 10.1890/08-0575.1. [DOI] [PubMed] [Google Scholar]
  17. Edburg SL, Hicke JA, Brooks PD, Pendall EG, Ewers BE, Norton U, Gochis D, Gutmann ED, Meddens AJH. Cascading impacts of bark beetle-caused tree mortality on coupled biogeophysical and biogeochemical processes. Front Ecol Environ. 2012;10:416–424. [Google Scholar]
  18. Edworthy AB, Drever MC, Martin K. Woodpeckers increase in abundance but maintain fecundity in response to an outbreak of mountain pine bark beetles. For Ecol Manag. 2011;261:203–210. [Google Scholar]
  19. Fischer A, Fischer HS, Kopecký M, Macek M, Wild J. Small changes in species composition despite stand-replacing bark beetle outbreak in Picea abies mountain forests. Can J For Res. 2015;45:1164–1171. [Google Scholar]
  20. Fontaine JB, Donato DC, Campbell JL, Martin JG, Law BE. Effects of post-fire logging on forest surface air temperatures in the Siskiyou Mountains, Oregon, USA. Forestry. 2010;83:477–482. [Google Scholar]
  21. González MA, Olea PP, Mateo-Tomás P, Garcia-Tejero S, De Frutos Á, Robles L, Purroy FJ, Ena V. Habitat selection and diet of Western Capercaillie Tetrao urogallus in an atypical biogeographical region. Ibis. 2012;154:260–272. [Google Scholar]
  22. Graf RF, Mathys L, Bollmann K. Habitat assessment for forest dwelling species using LiDAR remote sensing: capercaillie in the Alps. For Ecol Manag. 2009;257:160–167. [Google Scholar]
  23. Heurich M, Ochs T, Andresen T, Schneider T. Object-orientated image analysis for the semi-automatic detection of dead trees following a spruce bark beetle (Ips typographus) outbreak. Eur J For Res. 2010;129:313–324. [Google Scholar]
  24. Jönsson AM, Appelberg G, Harding S, Bärring L. Spatio-temporal impact of climate change on the activity and voltinism of the spruce bark beetle, Ips typographus. Glob Chang Biol. 2009;15:486–499. [Google Scholar]
  25. Kautz M, Dworschak K, Gruppe A, Schopf R. Quantifying spatio-temporal dispersion of bark beetle infestations in epidemic and non-epidemic conditions. For Ecol Manag. 2011;262:598–608. [Google Scholar]
  26. Klaus S. A 33-year study of hazel grouse Bonasa bonasia in the Bohemian Forest, Šumava, Czech Republic: effects of weather on density in autumn. Wildl Biol. 2007;13:105–108. [Google Scholar]
  27. Koivula M, Spence JR. Effects of post-fire salvage logging on boreal mixed-wood ground beetle assemblages (Coleoptera, Carabidae) For Ecol Manag. 2006;236:102–112. [Google Scholar]
  28. Kortmann M, Hurst J, Brinkmann R, Heurich M, Silveyra Gonzalez R, Müller J, Thorn S. Beauty and the beast: how a bat utilizes forests shaped by outbreaks of an insect pest. Anim Conserv. 2017:1–10. [Google Scholar]
  29. Kulakowski D, Seidl R, Holeksa J, Kuuluvainen T, Nagel TA, Panayotov M, Svoboda M, Thorn S, Vacchiano G, Whitlock C, Wohlgemuth T, et al. A walk on the wild side: disturbance dynamics and the conservation and management of European mountain forest ecosystems. For Ecol Manag. 2016:2–7. doi: 10.1016/j.foreco.2016.07.037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Latifi H, Heurich M, Hartig F, Müller J, Krzystek P, Jehl H, Dech S. Estimating over- and understorey canopy density of temperate mixed stands by airborne LiDAR data. Forestry. 2015;89:69–81. [Google Scholar]
  31. Latifi H, Hill S, Schumann B, Heurich M, Dech S. Multi-model estimation of understorey shrub, herb and moss cover in temperate forest stands by laser scanner data. Forestry. 2017;90:496–514. [Google Scholar]
  32. Lefcheck JS. PiecewiseSEM: piecewise structural equation modelling in R for ecology, evolution, and systematics. Methods Ecol Evol. 2016;7:573–579. [Google Scholar]
  33. Lehnert LW, Bassler C, Brandl R, Burton PJ, Müller J. Conservation value of forests attacked by bark beetles: highest number of indicator species is found in early successional stages. J Nat Conserv. 2013;21:97–104. [Google Scholar]
  34. Leverkus AB, Gustafsson L, María J, Benayas R, Castro J. Does post-disturbance salvage logging affect the provision of ecosystem services? A systematic review protocol. Environ Evid. 2015;4:16. [Google Scholar]
  35. Ludwig T, Klaus S. Habitat selection in the post-breeding period by Hazel Grouse Tetrastes bonasia in the Bohemian Forest. J Ornithol. 2016:1–12. [Google Scholar]
  36. Mikkelson KM, Bokman CM, Sharp O. Rare taxa maintain microbial diversity and contribute to terrestrial community dynamics throughout bark beetle infestation. Appl Environ Microbiol. 2016;82:6912–6919. doi: 10.1128/AEM.02245-16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Mikolás M, Svitok M, Bollmann K, Ba R, Janda P, Trotsiuk V, Vítková L, Teodosiu M, Coppes J, Schurman JS, Morrissey RC, et al. Mixed-severity natural disturbances promote the occurrence of an endangered umbrella species in primary forests. For Ecol Manag. 2017;405:210–218. [Google Scholar]
  38. Mitchell RG, Preisler HK. Fall rate of lodgepole pine killed by the mountain pine beetle in Central Oregon. West J Appl For. 1998;13:23–26. [Google Scholar]
  39. Müller D, Schröder B, Müller J. Modelling habitat selection of the cryptic Hazel Grouse Bonasa bonasia in a montane forest. J Ornithol. 2009;150:717–732. [Google Scholar]
  40. Næsset E. Practical large-scale forest stand inventory using a small-footprint airborne scanning laser. Scand J For Res. 2004;19:164–179. [Google Scholar]
  41. Naimi B. Usdm: Uncertainty Analysis for Species Distribution Models R Package Version 1.1–15. 2015 [Google Scholar]
  42. Paillet Y, Bergès L, Hjältèn J, Ódor P, Avon C, Bernhardt-Römermann M, Bijlsma RJ, De Bruyn L, Fuhr M, Grandin U, Kanka R. Biodiversity differences between managed and unmanaged forests: meta-analysis of species richness in Europe. Conserv Biol. 2010;24:101–112. doi: 10.1111/j.1523-1739.2009.01399.x. [DOI] [PubMed] [Google Scholar]
  43. Pausas JG, Parr CL. Towards an understanding of the evolutionary role of fire in animals. Evol Ecol. 2018:1–13. [Google Scholar]
  44. Pfandl B, Lentner R, Hochbichler E. Siedlungsdichten und Habitatpräferenzen ausge-wählter Waldvogelarten (Auerhuhn, Haselhuhn, Dreizehenspecht und Grauspecht) in den Brandenberger Alpen (Tirol) als Entscheidungshilfe für forstliche Planungen im Bergmischwald. Egretta. 2013;53:100–112. [Google Scholar]
  45. Pierson JC, Allendorf FW, Drapeau P, Schwartz MK. Breed locally, disperse globally: fine-scale genetic structure despite landscape-scale panmixia in a fire-specialist. PLoS ONE. 2013;8:e67248. doi: 10.1371/journal.pone.0067248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing; Vienna, Austria: 2013. [Google Scholar]
  47. Reitberger J, Schnörr C, Krzystek P, Stilla U. 3D segmentation of single trees exploiting full waveform LIDAR data. ISPRS J Photogramm. 2009;64:561–574. [Google Scholar]
  48. Rösner S, Brandl R, Segelbacher G, Lorenc T, Müller J. Noninvasive genetic sampling allows estimation of capercaillie numbers and population structure in the Bohemian Forest. Eur J Wildl Res. 2014a;60:789–801. [Google Scholar]
  49. Rösner S, Mussard-Forster E, Lorenc T, Müller J. Recreation shapes a "landscape of fear" for a threatened forest bird species in Central Europe. Landsc Ecol. 2014b;29:55–66. [Google Scholar]
  50. Rost J, Clavero M, Brotons L, Pons P. The effect of postfire salvage logging on bird communities in Mediterranean pine forests: the benefits for declining species. J Appl Ecol. 2012;49:644–651. [Google Scholar]
  51. Sachot S, Perrin N, Neet C. Winter habitat selection by two sympatric forest grouse in Switzerland: implications for conservation. Biol Conserv. 2003;112:373–382. [Google Scholar]
  52. Sahlsten J, Wickström F, Höglund J. Hazel grouse Bonasa bonasia population dynamics in a fragmented landscape: a metapopulation approach. Wildl Biol. 2010;16:35–46. [Google Scholar]
  53. Schäublin S, Bollmann K. Winter habitat selection and conservation of Hazel Grouse (Bonasa bonasia) in mountain forests. J Ornithol. 2011;152:179–192. [Google Scholar]
  54. Schelhaas MJ, Nabuurs GJ, Schuck A. Natural disturbances in the European forests in the 19th and 20th centuries. Glob Chang Biol. 2003;9:1620–1633. [Google Scholar]
  55. Seidl R, Rammer W. Climate change amplifies the interactions between wind and bark beetle disturbances in forest landscapes. Landsc Ecol. 2017:1–14. doi: 10.1007/s10980-016-0396-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Seidl R, Schelhaas M-J, Rammer W, Verkerk PJ. Increasing forest disturbances in Europe and their impact on carbon storage. Nat Clim Chang. 2014;4 doi: 10.1038/nclimate2318. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Seidl R, Müller J, Hothorn T, Bässler C, Heurich M, Kautz M. Small beetle, large-scale drivers: how regional and landscape factors affect outbreaks of the European spruce bark beetle. J Appl Ecol. 2016;53:530–540. doi: 10.1111/1365-2664.12540. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Selås V, Spidsø TK, Wiig Ø. Climate change in Norway: warm summers limit grouse reproduction. Popul Ecol. 2011;53:361–371. [Google Scholar]
  59. Senf C, Seidl R. Natural disturbances are spatially diverse but temporally synchronized across temperate forest landscapes in Europe. Glob Chang Biol. 2017:1–11. doi: 10.1111/gcb.13897. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Shipley B. Confirmatory path analysis in a generalized multilevel context. Ecology. 2009;90:363–368. doi: 10.1890/08-1034.1. [DOI] [PubMed] [Google Scholar]
  61. Shipley B. The AIC model selection method applied to path analytic models compared using a d-separation test. Ecology. 2013;94:560–564. doi: 10.1890/12-0976.1. [DOI] [PubMed] [Google Scholar]
  62. Storch I. Habitat selection by Capercaillie in summer and autumn - is bilberry important? Oecologia. 1993;95:257–265. doi: 10.1007/BF00323498. [DOI] [PubMed] [Google Scholar]
  63. Storch I. On spatial resolution in habitat models: can small-scale forest structure explain capercaillie numbers? Conserv Ecol. 2002;6:1–24. [Google Scholar]
  64. Storch I. Conservation status of grouse worldwide: an update. Wildl Biol. 2007a;13:5–12. [Google Scholar]
  65. Storch I. Grouse: Status Survey and Conservation Action Plan 2006-2010. 2007b [Google Scholar]
  66. Suter W, Graf RF, Hess R. Capercaillie (Tetrao urogallus) and avian biodiversity: testing the umbrella-species concept. Conserv Biol. 2002;16:778–788. [Google Scholar]
  67. Swanson ME, Franklin JF, Beschta RL, Crisafulli CM, DellaSala DA, Hutto RL, Lindenmayer DB, Swanson FJ. The forgotten stage of forest succession: early-successional ecosystems on forest sites. Front Ecol Environ. 2011;9:117–125. [Google Scholar]
  68. Teuscher M, Brandl R, Rösner S, Bufka L, Lorenc T, Förster B, Hothorn T, Müller J. Modelling habitat suitability for the Capercaillie Tetrao urogallus in the national parks Bavarian Forest and Šumava. Ornithol Anz. 2011;50:97–113. [Google Scholar]
  69. Thiel D, Unger C, Kéry M, Jenni L. Selection of night roosts in winter by Capercaillie Tetrao urogallus in Central Europe. Wildl Biol. 2007;13:73–86. [Google Scholar]
  70. Thom D, Rammer W, Seidl R. Disturbances catalyze the adaptation of forest ecosystems to changing climate conditions. Glob Chang Biol. 2017;23:269–282. doi: 10.1111/gcb.13506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Thorn S, Bässler C, Bußler H, Lindenmayer DB, Schmidt S, Seibold S, Wende B, Müller J. Bark-scratching of storm-felled trees preserves biodiversity at lower economic costs compared to debarking. For Ecol Manag. 2016;364:10–16. [Google Scholar]
  72. Vauhkonen J, Imponen J. Unsupervised classification of airborne laser scanning data to locate potential wildlife habitats for forest management planning. Forestry. 2016;89:350–363. [Google Scholar]
  73. Wermelinger B. Ecology and management of the spruce bark beetle Ips typographus - a review of recent research. For Ecol Manag. 2004;202:67–82. [Google Scholar]
  74. Wild J, Kopecky M, Svoboda M, Zenahlikova M, Edwards-Jonasova M, Herben T. Spatial patterns with memory: tree regeneration after stand-replacing disturbance in Picea abies mountain forests. J Veg Sci. 2014:1–14. [Google Scholar]
  75. Yao W, Krzystek P, Heurich M. Tree species classification and estimation of stem volume and DBH based on single tree extraction by exploiting airborne full-waveform LiDAR data Remote Sens. Environ. 2012;123:368–380. [Google Scholar]
  76. Yao W, Krull J, Krzystek P, Heurich M. Sensitivity analysis of 3D individual tree detection from LiDAR point clouds of temperate forests. Forests. 2014;5:1122–1142. [Google Scholar]
  77. Zellweger F, Braunisch V, Baltensweiler A, Bollmann K. Remotely sensed forest structural complexity predicts multi species occurrence at the landscape scale. For Ecol Manag. 2013;307:303–312. doi: 10.1016/j.foreco.2013.07.023. [DOI] [Google Scholar]
  78. Zellweger F, Morsdorf F, Purves RS, Braunisch V, Bollmann K. Improved methods for measuring forest landscape structure: LiDAR complements field-based habitat assessment. Biodivers Conserv. 2014;23:289–307. [Google Scholar]
  79. Zeppenfeld T, Svoboda M, DeRose RJ, Heurich M, Müller J, Cizkova P, Stary M, Bace R, Donato DC. Response of mountain Picea abies forests to stand-replacing bark beetle outbreaks: neighbourhood effects lead to self-replacement. J Appl Ecol. 2015;52:1402–1411. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Appendix
EMS145170-supplement-Appendix.docx (100.2KB, docx)

RESOURCES