Long-term surveys of ungulates’ effects on tree and shrub species in mountainous forests – a valuable outcome or a biased picture?

doi:10.21203/rs.3.rs-4842616/v1

Download PDF

Research Article

Long-term surveys of ungulates’ effects on tree and shrub species in mountainous forests – a valuable outcome or a biased picture?

https://doi.org/10.21203/rs.3.rs-4842616/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Ungulate herbivory might induce different effects on the diversity and growth of trees and shrubs. The density, distribution, and the species of ungulates as well as plant communities’ composition and other factors determine whether ungulate herbivory promotes or limits plants’ diversity and growth. The impacts of ungulates on woody plants is commonly surveyed with exclosure-control approaches. In practice, such surveys frequently only cover short periods of time, addressing immediate management needs. Long-term surveys, documenting lasting effects of ungulate herbivory, are highly needed, but still rare. However, the general transferability of outcomes of long-term surveys might be limited due different disturbing factors. This study addresses two basic aspects of long-term monitoring in mountainous forests, based on a unique 30-year data set: (1) Possible long-term effects of herbivores on forest vegetation (e.g., species/structural diversity of woody plants) and (2) potential biases of a long-term survey. In our study, diversity of woody plant species showed great variability with no significant impact of ungulate herbivory. The presence of ungulates had a significantly negative effect on the vertical structural diversity and growth of trees. Due to the slower growth on control plots, it took trees longer to reach a 160 cm height-threshold with their terminal shoots than on exclosure plots. Our long-term control-exclosure data set indeed reflected a potential bias towards specific site conditions. This included several factors, like site-specific growth patterns of trees, occurrences of natural abiotic disturbances that influence the functional life of exclosures, and others.

Herbivory

woody plants

diversity indices

hill numbers

exclosure-control pairs

height growth

Herbivory, defined as the consumption of living plant tissues by animals, comprises a number of different processes like seed predation, browsing, grazing, mining, or bark peeling (Schowalter 2016). It is one of the most crucial disturbances within an individual plant’s ontogeny (Aguirrebengoa et al. 2018; Massad 2013; Selaković et al. 2017) and has major impacts on species composition and biodiversity of plant communities and ecosystems (Bernes et al. 2018). Starting with seedling herbivory, herbivores shape landscapes as their preference for certain seedlings leads to cascading changes in plant species communities, biodiversity, and forest regeneration (Cilles et al. 2016; Williams and Brodie 2023). The effects of leaf browsing and grazing on plants range from differences in morphology (Persson et al. 2005) to changes in metabolism (Gaquerel et al. 2014; Schrijvers-Gonlag et al. 2020; Zhou et al. 2015) and alterations in interspecies competition (Sebata 2013). These small-scale effects can in turn lead to changes in species compositions of plants and animals (Wisdom et al. 2006), nutrient availability, and nutrient cycles (Abbas et al. 2012; Murray et al. 2013).

In temperate forest ecosystems, ungulates are one of the most impactful herbivores (VanderMolen and Webster 2021; Weisberg and Bugmann 2003). Ungulates prefer certain plant species or plant individuals to others as their availability, accessibility, nutrient value, defense status and many other factors can differ significantly (Holeski et al. 2016; Ingram et al. 2013). The selective feeding by ungulates can tip the scales in the inter-species competition of plants to favor one side or the other (Maxwell et al. 2019), as herbivory can lead to a complete disappearance of seedlings and smaller plants or reduced height growth of saplings and bigger individuals (Beguin et al. 2016; Hood and Bayley 2009). These effects can be exacerbated through inherently slow annual height growth due to environmental factors such as soil fertility (Zamora et al. 2001). Selective feeding can lead to changes in species composition, productivity, and biodiversity of plant communities (Royo and Carson 2022), which in turn might result in a change in animal communities (Seager et al. 2013). The extent of related changes also depends on environmental drivers (e.g., aspect or light availability), potentially reducing or amplifying impacts of ungulates on the vegetation (Kupferschmid et al. 2014, 2020).

Although wild ruminants are constitutive elements of many terrestrial ecosystems, their impact on vegetation can be assessed negatively from a human perspective if changes in spatio-temporal habitat use patterns, unnaturally high ruminant densities or local spatial concentration effects occur and threaten human interests (Reimoser et al. 1999, 2023; Reimoser 2003). Even in protected areas, herbivore influences can be assessed negatively from a human perspective if these influences hinder the achievement of previously defined conservation targets (Reimoser et al. 2022). This can also be the case in areas that target at process protection, e.g., when hunting is suspended in the protected area resulting in a high settling stimulus for wild ruminants, or when an area represents a more attractive habitat than the surrounding landscape for other reasons. In such constellations, excessive ruminant densities might occur, which may jeopardize conservation objectives. It is therefore highly relevant in managed forests, but also in forests in protected areas, to monitor wild ruminant impacts and, if necessary, to derive management needs and appropriate measures.

The impacts of ungulate herbivory on trees have been the subject of many surveys, with exclosure-control settings being the most common approach (Bellingham and Allan 2003; Casabon and Pothier 2007; Millett and Edmondson 2013; Pellerin et al. 2010; Reimoser et al. 2023; Russell and Fowler 2004; Tschöpe et al. 2011). By comparing vegetation data from a fenced plot to which the ungulates have no access, with vegetation data from an accessible control plot ungulates’ impacts can be separated from other driving factors. This allows for conclusions on the effect of ungulates on biodiversity, height growth, forest regeneration, species composition, and more (Hedwall et al. 2018; Kay and Bartos 2000; McGarvey et al. 2013; Perkovich and Ward 2022; Seymour et al. 2016).

As the most vulnerable part of a tree’s life cycle for ungulate herbivory lies in its early years, many exclosure studies particularly address these early stages and related community responses. However, the complexity of plant-animal interactions together with intra- and interspecific interactions between plants/species makes it difficult to translate recent influences of wild ruminants into impacts at later developmental stages. Compensatory mortalities, gap-forming and gap-filling dynamics, human intervention as well as interactions with abiotic drivers might play a decisive role. It has been shown that early ungulate herbivory has diverse effects on biodiversity (Nopp-Mayr et al. 2023) and height growth patterns (Nopp-Mayr et al. 2020) in later years and might be superimposed by responses to other environmental drivers. As a consequence, lasting effects of ungulate herbivory might be overestimated (Nopp-Mayr et al. 2020), when only covering the first years after germination and recording all tree individuals (total stem density) within monitoring schemes. In practice, surveys of herbivore impacts frequently only cover short periods of time, addressing immediate management needs, but largely ignore further-reaching effects that cannot be assessed within these windows of time. Thus, long-term surveys, documenting lasting effects of ungulate herbivory are needed, but still rare.

However, general transferability of outcomes of long-term surveys might not be automatically assumed, but potential representativeness only for specific conditions that allow for such a monitoring over decades might be given. This issue is of general relevance when monitoring ungulates’ impacts on forest vegetation on a long-term basis.

In the present study, we could address these two basic aspects of long-term monitoring in mountainous forests: (1) Possible long-term effects of herbivores on forest vegetation (i.e., on the species diversity and structural diversity of woody plants and on height growth patterns) and (2) possible biases of a long-term survey (by generating subsamples with different survey durations). Our study is based on a regular state-wide ungulate herbivory monitoring system in the westernmost state of Austria (Vorarlberg), which provides a unique data set of 30-year surveys. This is the longest lasting statewide ungulate impact survey in Central European mountainous forests with a large number of sample plots. Within this system ongoing ungulate impacts and their lasting effects on forest regeneration are monitored and evaluated, based on comparisons of ungulate exclosure plot data with data from open control plots accessible for ungulates. There are plot pairs, which have been recorded for the last 30 years having their fence intact the whole time.

The survey allowed us to explore potential advantages of long-term data sets, capturing the complex cascades of competition, compensatory mortality and more at later developmental stages, which would have been missed in the course of short- or mid-term surveys. This includes the following questions being relevant for management issues: (1) Are effects of ungulate herbivory in terms of tree or shrub species diversity detectable after a 30-year time span? (2) Is the structural diversity (in terms of height class diversity of trees) different in the fenced plots compared to the control plots after 30 years? (3) Do plot types (open access vs. exclosure) differ in height class distribution of trees on the long term? (4) Does the number of years till terminal shoots of trees outgrow the height of ungulate browsing differ between the fenced plots and the control plots? On the other hand, dividing the data set into subsets covering different windows of time allowed us to (5) test the extent to which long-term data are biased towards slow-growing sites.

Study area

Forest regeneration was monitored in the westernmost state of Austria (Vorarlberg) with a proportion of forest cover of 36% (Fig. 1). The state’s geology changes distinctly along a north-south gradient from the Molasse and Helvetic sediment zones to a Flysch zone, then the Northern Limestone Alps, and in the south the crystalline rocks of the Central Eastern Alps. The annual mean temperature in the study area ranges between 6 and 9°C and the mean annual precipitation in the surveyed areas lies approximately between 1.200 to 2.400 mm. The monthly precipitation pattern shows a maximum in June although in some areas a second maximum in November is typical. The duration of snow cover varies strongly between survey plot pairs with 50 days for plots closer to the valley bottoms to 180 days in areas around 1.500 m above sea level. The elevation range, covered by plot pairs, reaches from 500 m a.s.l. to approximately 1.700 m a.s.l..

The plot pairs are assigned to different representative forest communities: (i) Submontane deciduous forests with dominating European beech (Fagus sylvatica), frequent European ash (Fraxinus excelsior) and sycamore maple (Acer pseuodoplatanus); Norway spruce (Picea abies) and silver fir (Abies alba) often promoted through silviculture. (ii) Montane mixed forest communities consisting mainly of Norway spruce, silver fir, and European beech, with European ash, sycamore maple, and wych elm (Ulmus glabra) mixed in. (iii) Upper montane forests (mainly Norway spruce and silver fir). (iv) Subalpine Norway spruce forests at higher altitudes with Mountain ash (Sorbus aucuparia) mixed in. In Vorarlberg, most forests are managed with silvicultural systems aiming at permanent forests without clearcuts, as approximately 50% of the state’s forests are protection forests against avalanches, rock falls, and erosion. The most common type of permanent forest is the selection forest.

Ungulate densities are high, with the most common ungulate species being red deer (Cervus elaphus), roe deer (Capreolus capreolus), and chamois (Rupicapra rupicapra). The state tries to keep the flat terrain in its northern and north-western parts free of red deer. Yearly hunting bags are on average 1 red deer, 2 roe deer, and 1 chamois per km² (based on a hunting season from May to December, see also Table 1).

Table 1

Average yearly hunting bags (individuals/km²) of the main three ungulate species within two different time periods. Minimum and maximum hunting bags within the entire time horizon (1985–2022) are given for the respective years
Period or year	Hunting bag (n/km²)
	Red deer	Roe deer	Chamois
1985–1994	0.7	1.7	0.6
2015–2022	1.1	2.0	0.5
Min. year	0.5 (1990)	1.5 (1989/90)	0.3 (2007)
Max. year	1.3 (2013)	2.3 (2010/12)	0.7 (1987)

Field Sampling

With an amendment in Vorarlberg’s hunting regulations in 1988, the calculation basis for the annual game harvest plan changed from imprecise ungulate counts to the extent of damages caused by ungulates. To monitor ungulate damage, pairs of fenced plots and unfenced control plots were established in the state’s forests and were to be surveyed every 3 years. In total 3405 pairs of ungulate exclosures and unfenced control plots were established at altitudes between 500–1.100 m a.s.l. since 1989, the last survey was in 2022. Until this point of time, many fence or control plots had been destroyed or disturbed by natural abiotic disturbance events like storm throw, snow loading, or rock falls. Plots that experienced other disturbances than ungulate herbivory were omitted. Thus, a final number of 73 plot pairs could be surveyed over the entire 30-year time span.

The first survey (= time of plot establishment, year 0) and the second survey (year 3) were carried out by F. Reimoser and team, the last survey 2022 by W. Lechner and F. Reimoser. The years in between were sampled by forest supervisors and many plot pairs had been (unplanned) abandoned by them after several surveys, when a certain tree height was reached.

The guidelines of the state Vorarlberg for the monitoring of ungulate impacts/damages via plot pairs define bindingly how and where plot pairs are to be created (see also (Reimoser et al. 2023)). They must be installed on areas where forest regeneration is in an early stage (saplings beneath 50 cm in height) or is expected to start soon. Canopy cover needs to be beneath a threshold (ca. 80%) that allows for development of seedlings on the ground after germination depending on forest community and altitude.

The plot pairs contained only natural forest regeneration and no planted trees. The forests in which the plot pairs are located are all subject to forest management. The vegetation within the plots however was not tampered with. Plots that had been directly affected by forest management were not sampled. If the regeneration of a plot pair was influenced unequally between a fenced and paired unfenced plot by forest operations in the mature stand, this plot pair was omitted from the investigation. The entire sampling procedure was optimized to capture effects of ungulate herbivory and to keep other driving factors as constant as possible.

The fenced plots are designed to keep out larger (ungulate) herbivores (i.e., red deer, roe deer and chamois). Smaller herbivores like the European hare (Lepus europaeus) or rodents still have access to the fenced plots. To ensure comparability within a plot pair, plot sites are chosen with highest possible similarity in environmental conditions. Which of the two plots is fenced is determined at random by a coin toss. Exclosure plots are surrounded by a 2 m tall mesh fence, which encloses an area of 6 x 6 m. Plots are numbered and marked with wooden stakes in the middle and the corners. The fenced plot and the control plot are located at distances between 5 and 40 m. Larger distances would reduce comparability within a pair, whereas limiting the distance to a shorter range than approximately 40 m would increase the difficulty of finding two suitable and similar spots, which is especially challenging in steep and uneven alpine terrain. To eliminate edge effects, we surveyed a 5 x 5 m area within the 6 x 6 m zone of each exclosure and control plot.

Following the gap-maker–gap-filler approach (Lertzman 1992), we dynamically recorded the height of the largest tree individuals per tree species on the paired plots (open access and ungulate exclosure, each 25 m²), as this allowed us to assess herbivore impacts on those individuals, which have a high probability of attaining later developmental stages (Nopp-Mayr et al. 2020). We surveyed the six highest individuals per tree species and plot or, in case of lower stem numbers (< 6), all occurring individuals per tree species and sampling plot (see Nopp-Mayr et al. 2020). We used the following height classes: 10 cm, 11–25 cm, 26–40 cm, 41–70 cm, 71–100 cm, 101–130 cm, 131–160 cm, 161–200 cm, 201–250 cm, 251–300 cm, … (continuing 50 cm classes). It has to be noted that the survey scheme per se does not provide for a termination of the surveys when a certain tree height (e.g. 2 m) is reached. The height classes cover an increasing height range with increasing height, starting with 15 cm intervals for seedlings and ending up with 50 cm intervals at later stages. These unequal class ranges were chosen for two reasons: (1) They account for typical height growth patterns in tree species with steep slopes at early life stages and flattening curves after decades (Manso et al. 2021); (2) consequently, small differences in height or height growth might be more meaningful in terms of competitive power and survival in smaller trees compared to later stages (Adams et al. 2007; Matisons et al. 2020; Walters et al. 2020). For shrubs, we visually assessed the coverage (%) of shrub species per survey plot, as individual entities and related numbers might not be easily identified in the course of a field survey.

No animal experiments were carried out. The ungulates were neither caught nor influenced by medication. They were only denied access to small patches within forests by means of a fence. The fencing was carried out in accordance with all relevant guidelines and regulations.

Data preparation

We used the occurrence of different height classes to analyze long-term differences in the structural diversity of forest regeneration induced by ungulate herbivory. As this does not represent actual differences in absolute height growth between the exclosure and the control plots we further addressed height differences of the highest 6 individuals per tree species. To explore differences in diversity of shrub species, coverage data per shrub species was used as representative of species occurrence.

To address the question whether ungulates impact the time a tree species needs to grow out of the ungulate’s feeding range, we focused on the highest tree per tree species recorded in a 3-year cycle. The height, at which ungulates are assumed to have little to no impact on tree height growth was set to 160 cm. This height was chosen as red deer, the tallest of the three occurring ungulate species, very rarely feed on vegetation above this height (Renaud et al. 2003). To highlight effects of short-term vs. long-term surveys in terms of height growth performance, we created different data subsets covering three different windows of time (a: 9 years; b: 15 years; c: 24 years; see also Table 2). Per subset we calculated the share of plots, where a tree species reached a height of 160 cm in percent of the total number of surveyed plots. These percentages were contrasted between the open access control plots and the exclosures.

Table 2

Data sets, sampling periods, sample sizes and related analyses of the plot pairs (open control vs. ungulate exclosures)
Sampling time/ period (years)	n (plot pairs)	Analyses (control vs. exclosures)
30	73	Tree species diversity, height class diversity of trees, height class distribution of trees, shrub species diversity
9	1722	Share of plots, where tree species reached a height of 160 cm in percent of the total number of plots (n)
15	907	Share of plots, where tree species reached a height of 160 cm in percent of the total number of plots (n)
24	332	Share of plots, where tree species reached a height of 160 cm in percent of the total number of plots (n)

Statistical analyses

According to Chao et al. (2020), woody plants of the same species tend to aggregate spatially, potentially violating the assumption of individuals being independent sampling units. Therefore, the abundance data of trees and shrubs were transformed into incidence-based data, which, in a species-by-sampling-unit incidence matrix, can be used for incidence-based diversity analyses. Accordingly, height data of the six tallest tree individuals per species and plot was transformed into an incidence-based height class dataset. This means that each plot is handled as an independent sampling unit (Chao et al. 2014; Chao et al. 2020).

To compare the impacts of ungulates on the diversity of woody plants and height classes, diversity estimates are needed. The first three Hill numbers of order q were chosen as these are commonly used diversity estimates. Hill number q = 0 refers to species richness and is henceforth referred to as “species richness”. Hill number q = 1, the exponential of Shannon’s entropy index, is henceforth referred to as “Shannon diversity” and q = 2, the inverse Simpson´s concentration index, is henceforth referred to as “Simpson diversity” (Chao et al. 2014). A major difference between the three mentioned Hill numbers is how much weight is attributed to rare species. Species richness is calculated with equal weight for rare species and common species because their detection probabilities are not accounted for. Shannon diversity on the other hand includes the detection probability and Simpson diversity goes even further and incorporates the squared detection probability. Therefore, species richness is best suited to address the diversity of rare species, Shannon diversity is best suited to address the diversity of frequent species and Simpson diversity is best suited to address the diversity of highly frequent species. In Chao et al. (2014, 2020) and Hill (1973) a more in-depth explanation is provided.

The calculation of the three chosen diversity measures followed the description by Colwell et al. (2012), Chao et al. (2020), and Nopp-Mayr et al. (2023). The first step was the calculation of sample size-based rarefaction and extrapolation curves. The important role sample size plays in the detection and occurrence of species were the reason sample size-based curves were chosen. The extrapolation was used to double the sample size of surveyed fenced and control plots (Chao et al. 2014; Nopp-Mayr et al. 2023).(Chao et al. (2020) state that “true” diversity is observed if the sample size-based rarefaction and extrapolation curve approaches an asymptote. If this is not the case a comparison of diversity between the exclosure and control plots is not possible with this analysis. After calculating the curves, we assured that the data is sufficient. This was done via a visual examination of sample completeness profiles (Chao et al. 2020). As described in Nopp-Mayr et al. (2023) sample size-based rarefaction and extrapolation curves for Hill numbers of order q = 1 (Shannon diversity) and q = 2 (Simpson diversity) usually level off. This indicates that an interpretation of frequent and highly common species as well as a comparison between the assemblages is possible. The opposite is the case for Species richness (q = 0), for which in most cases the sample size-based rarefaction and extrapolation curves do not approach an asymptote. This means that the sample size is insufficient to interpret the rare species in an assemblage and the asymptotic estimator can only be interpreted as a lower bound of the true species diversity. To still be able to interpret such cases, a non-asymptotic standardization approach was utilized (Chao et al. 2020). The calculation of diversity estimates based on standardized sample coverage (maximum coverage C_max) allowed for a comparison between assemblages, even when the sample size-based rarefaction and extrapolation curves did not approach an asymptote (Chao et al. 2020). To determine whether significant differences between exclosure and control plots existed 84% confidence intervals were calculated. If the confidence intervals do not overlap a significant difference between the two groups is detected. Payton et al. (2003) found that determining significant differences via the overlapping or non-overlapping of 95% confidence intervals is too conservative. They recommend 84% confidence intervals for the comparison of two samples with approximately equal standard errors (see also by Colwell et al. 2012). To define the number of bootstrap replications needed for the diversity analyses different amounts of replications were experimented with. At 1000 replications the calculated estimates stabilized, and this number of replications was henceforth used.

Additionally, we ran a Wilcoxon signed-rank test to compare the height of the tallest individual per tree species and plot type (open access vs. exclosure).

To calculate the diversity measures the R packages “iNext” (Hsieh et al. 2016) and “iNext.4step” (Chao et al. 2020) were used. Graphs were created with the R package “ggplot2” (Wilkinson 2011).

Tree species diversity

During the starting survey (year 0) a total of 18 different tree species were recorded on the 73 plot pairs that were monitored over 30 years (see Table 3). The survey in year 30 yielded 14 different tree species on the control plots and 13 on the exclosure plots. European larch (Larix decidua), linden (Tilia sp.), Scots pine (Pinus sylvestris), and willow (Salix sp.) were only detected during the starting survey. In year 30, common walnut (Juglans regia) and oak (Quercus sp.) were exclusively detected on a single control plot, whereas common yew was exclusively detected on exclosure plots. The minimum number of trees species detected on a single plot was 1 regardless of the treatment and the time of recording. The maximum number of species detected on a single plot however was different between years and treatment. On control plots, a maximum of 12 tree species in year 0 and 8 species in year 30 were detected (Table 3). In comparison, a maximum of 11 and 9 tree species was recorded on exclosure plots in year 0 and year 30, respectively. A list containing all detected species is available in the Appendix (Table A1).

Table 3

Total number of tree species detected on the 73 control and exclosure plots in the starting (year 0) and last survey (year 30). n = total number of species found on the 73 plots, min = minimum number of species detected on a single plot, max = maximum number of species detected on a single plot, median = median number of species detected on a single plot
	Control plots
year	min	max	median	n	min	max	median	n
0	1	12	3	18	1	11	4	18
30	1	8	4	14	1	9	4	13

Sample size-based rarefaction and extrapolation sampling curves for Shannon diversity (q = 1) and Simpson diversity (q = 2) levelled off and approached an asymptote (see Fig. 2) for both control and exclosure plots in years 0 and 30. This means that the sample size was sufficient to draw conclusions on frequent and highly frequent species with this method and the curves could be interpreted without additional measures. The opposite was the case for species richness (q = 0). The curves did not approach an asymptote and therefore only represented a lower bound. To infer “true” species richness non-asymptotic point diversity estimates had to be calculated at a standardized coverage value of C_max = 99.7% (see Table 4).

Table 4

Asymptotic and non-asymptotic **tree species diversity** estimates for **year 0** and **year 30**. Non-asymptotic rarefaction and extrapolation analyses are based on point diversities for a specified sample coverage of C_max = 99.7%. The Hill numbers of order q stand for species richness (q = 0), Shannon diversity (q = 1), and Simpson diversity (q = 2)
	Control Plots			Exclosure Plots
Year	q = 0	q = 1	q = 2	q = 0	q = 1	q = 2
	Asymptotic diversity estimate (mean ± SE)
0	19.97 ± 5.34	8.55 ± 0.48	6.56 ± 0.28	20.63 ± 6.14	8.84 ± 0.47	6.89 ± 0.27
30	16.95 ± 4.06	7.77 ± 0.36	6.45 ± 0.21	13.98 ± 1.59	7.62 ± 0.33	6.36 ± 0.25
	Non-asymptotic point diversity estimate [84% confidence interval: lower; upper]
0	19.57 [12.8; 26.2]	8.42 [7.7; 9.0]	6.52 [6.1; 6.9]	20.04 [12.8; 27.2]	8.74 [8.1; 9.3]	6.86 [6.4; 7.2]
30	15.87 [11.5; 20.1]	7.70 [7.2; 8.1]	6.42 [6.1; 6.7]	13.4 [11.6; 15.2]	7.52 [7.0; 7.9]	6.32 [6.0; 6.6]

For both control and fenced plots, the starting survey (year 0) yielded higher values for all three diversity measures than the last survey (year 30). Regardless of year, the 84% confidence intervals overlapped for all the diversity measures indicating no significant differences in tree species diversity between control plots and fenced plots.

Height class diversity of trees

On the 73 plot pairs, surveyed for 30 years, a total of 29 different height classes were recorded, of which 6 were only encountered on exclosure plots. A minimum of 2 and a maximum of 15 height classes were recorded on control plots, whereas a minimum of 3 and a maximum of 16 height classes were recorded on exclosure plots. Control plots yielded a median of 6.5 height classes per plot and exclosure plots a median of 8. Sample size-based rarefaction and extrapolation sampling curves for Shannon diversity (q = 1) and Simpson diversity (q = 2) approached an asymptote (see Fig. 3) for both control and exclosure plots. This means that the sample size was sufficient to draw conclusions on frequent and highly frequent height classes with this method and the curves could be interpreted without additional measures. This did not apply for height class richness as the sample size-based rarefaction and extrapolation sampling curves did not level off. Therefore, the sample size was not large enough to infer “true” height class richness and the curves only represented a lower bound. To interpret height class richness non-asymptotic point diversity estimates at a standardized coverage value of C_max = 99.4% were calculated (see Table 5).

Table 5

Asymptotic and non-asymptotic **height class diversity** estimates of the **tallest individuals** (n_max = 6) per tree species **30 years** after establishment. Non-asymptotic rarefaction and extrapolation analyses are based on point diversities for a specified sample of C_max = 99.4%. The Hill numbers of order q stand for height class richness (q = 0), Shannon diversity (q = 1), and Simpson diversity (q = 2)
Control Plots			Exclosure Plots
q = 0	q = 1	q = 2	q = 0	q = 1	q = 2
Asymptotic diversity estimate (mean ± SE)
25.95 ± 2.41	15.11 ± 0.46	12.24 ± 0.39	41.32 ± 10.96	20.83 ± 0.60	17.98 ± 0.42
Non-asymptotic point diversity estimate [84% confidence interval: lower; upper]
23.29 [20.6; 25.9]	14.81 [14.1; 15.4]	12.10 [11.5; 12.6]	33.10 [23.8; 42.3]	20.57 [19.6; 21.4]	17.83 [17.1; 18.5]

Both Shannon diversity (q = 1) and Simpson diversity (q = 2) differed significantly between control and exclosure plots as indicated by non-overlapping confidence intervals. Frequent height classes were represented by a Shannon diversity of 20.57 in exclosure plots and 14.81 in control plots. The differences in Simpson diversity (17.83 in exclosure plots versus 12.10 in control plots) illustrated the impact of ungulate herbivory on highly frequent height classes. No significant differences in rare height classes between control and exclosure plots could be discovered as the confidence intervals of the height class richness overlapped.

Height class distribution

As differences in height class diversity do not inform about actual height growth differences between control and exclosure plots, we additionally contrasted the heights of the tallest individual per tree species and plot type (open access vs. exclosure) (Fig. 4). We used the mid-point height per height class interval as height representative for each tree (e.g. 85 cm for the 71–100 cm height class). A two-samples Wilcoxon test yielded significant height differences between control and exclosure plots in year 30 (p < 0.001) when pooling all tree species, reflecting generally suppressed height growth on the control plots. This delay can be observed in many, but not all tree species (Fig. 5), with Norway spruce being a clear exception.

Shrub species diversity

We recorded 19 different shrub species on 59 plot pairs in the course of the starting survey (year 0; Table 6). On some plot pairs, no shrubs were detected. Within these 19 species, English holly (Ilex aquifolium) was only detected on a single control plot and a Rhamnus species was only detected on a single exclosure plot.

Table 6

List of observed shrub species (in alphabetical order) on the control plots (C) and within the exclosures (E) in the year of establishment (year 0) and in year 30 after establishment
Shrub species	Year 0		Year 30
	C	E	C	E
Alder buckthorn (Rhamnus frangula)	x	x	---	x
Barberry (Berberis vulgaris)	x	x	---	---
Bilberry (Vaccinium myrtillus)	x	x	x	x
Black elder (Sambucus nigra)	x	x	x	---
Blackberry (Rubus fruticosus)	x	x	x	x
Buckthorn (Rhamnus sp.)	---	x	---	---
Common dogwood (Cornus sanguinea)	x	x	---	x
Common heather (Calluna vulgaris)	x	x	---	---
Daphne (Daphne sp.)	x	x	x	x
English holly (Ilex aquifolium)	x	---	x	x
English ivy (Hedera helix)	x	x	x	x
Hazel (Corylus avellana)	x	x	x	x
Hedge cherry (Lonicera sp.)	x	x	x	x
Lingonberry (Vaccinium vitis-idaea)	---	---	---	x
Privet (Ligustrum vulgare)	x	x	x	x
Raspberry (Rubus idaeus)	x	x	x	x
Red elder (Sambucus racemosa)	x	x	x	x
Rose (Rosa sp.)	x	x	x	x
Spindle bush (Euonymus europaeus)	x	x	---	---
Wooly snowball (Viburnum lantana)	x	x	x	x

In year 30, all 73 plot pairs were included in the analysis and a total of 17 different shrub species were recorded. European black elderberry (Sambucus nigra) was only detected on a single control plot. Lingonberry (Vaccinium vitis-idaea), common heather (Calluna vulgaris), common dogwood (Cornus sanguinea), and alder buckthorn (Rhamnus frangula) were only detected in exclosure plots. During the starting survey between 1 and 6 different shrub species were detected on control plots and between 1 and 5 on exclosure plots. Control and exclosure plots both had a median of 2 shrub species. After 30 years, between 1 and 4 shrub species were detected on control plots and between 1 and 5 on exclosure plots. The median for control plots changed to 1.5 shrub species for control plots and stayed at 2 shrub species for exclosure plots.

Sample size-based rarefaction and extrapolation sampling curves of Simpson diversity (q = 2) levelled off and approached an asymptote (see Fig. 6) for both control and exclosure plots in years 0 and 30. Shannon diversity (q = 1) only levelled off for the 30-year data. This means that the sample size was sufficient to draw conclusions on frequent (only year 30) and highly frequent species with this method and the curves could be interpreted without additional measures. The opposite was the case for Shannon diversity (q = 1) of year 0 and species richness (q = 0) of both time points. The curves did not approach an asymptote and therefore only represented a lower bound. To infer “true” species richness and Shannon diversity (only for year 0) non-asymptotic point diversity estimates had to be calculated at a standardized coverage value of C_max = 92.1% (see Table 6).

Table 6

Asymptotic and non-asymptotic **shrub diversity** estimates for **year 0** and **year 30**. Non-asymptotic rarefaction and extrapolation analyses are based on point diversities for a specified sample coverage of C_max = 92.1%. The Hill numbers of order q stand for species richness (q = 0), Shannon diversity (q = 1), and Simpson diversity (q = 2)
	Control Plots			Exclosure Plots
Year	q = 0	q = 1	q = 2	q = 0	q = 1	q = 2
Asymptotic diversity estimate (mean ± SE)
0	67.15 ± 33.03	10.36 ± 1.91	5.80 ± 0.61	57.81 ± 25.26	11.01 ± 1.59	6.85 ± 0.66
30	17.10 ± 5.85	6.58 ± 0.62	4.63 ± 0.41	20.10 ± 8.60	8.42 ± 0.75	5.47 ± 0.52
Non-asymptotic point diversity estimate [84% confidence interval: lower; upper]
0	27.03 [1.7; 52.2]	9.07 [6.3; 11.7]	5.69 [4.8; 6.5]	18.53 [2.0; 35.0]	9.14 [6.7; 11.5]	6.58 [5.5; 7.5]
30	15.87 [7.7; 24.0]	6.39 [5.4; 7.3]	4.57 [4.0; 5.1]	18.7 [10.2; 27.1]	8.15 [7.0; 9.2]	5.39 [4.6; 6.0]

No significant differences in shrub species diversity were detected between control and exclosure plots in year 0 and year 30 as the confidence intervals overlapped for all diversity measures.

Temporal pattern of the proportion of plots where terminal shoots outgrow the ungulate browsing height

Control plots vs. exclosures

To highlight potential differences between control and exclosure plots in the time the trees needed to outgrow ungulate browsing height, we used three subsamples of plot pairs that were surveyed in 3-year intervals for 9, 15 and 24 years. We defined height class 8 (> 160 cm) as related threshold, where trees´ terminal shoots are usually too high above ground to be accessed by ungulates under average conditions (i.e., outside periods with high and walkable snow coverage). In many cases, open access plots showed lower proportions and time delays till trees reached the 160 cm height threshold (Table 7). The delay was especially apparent in wild cherry (Prunus avium) and wych elm (Ulmus glabra). European beech (Fagus sylvatica), silver birch (Betula pendula), mountain ash (Sorbus aucuparia), and others showed distinctly lower percentages of plots where trees reached 160 cm tree heights. Within the frequently occurring tree species, the smallest differences between control and exclosure plots were observed for Norway spruce (Picea abies). It has to be mentioned that most tree species never reached height class 8 on most sites. This is also the reason why the percentages do not add up to 100.

Short-term vs. long-term survey duration

To address the question, whether long-term surveys might be biased in terms of growth performance, we calculated the share of plots, where the occurring tree species reached a height of 160 cm in percent of the total number of surveyed plots, respectively. These percentages were depicted for three different windows of time (a: 9 years; b: 15 years; c: 24 years). It became evident, that the portion of plots, where trees reached the height threshold, showed an earlier peak within the collective of plot pairs being surveyed for shorter periods of time (Table 7). This was the case both for open access and exclosure plots and also, when not particularly addressing a specific tree species, but any occurring tree species (see Table 7, column “Overall”). These observations suggest that long-term survey data indeed carry a bias in terms of height growth pattern. This means that lasting effects of ungulate impacts are reflected for specific situations, but not but not necessarily for the entire area of interest where monitoring takes place.

Table 7

Share of plots, where tree species reached a height of 160 cm in percent of the total number of plots (n). n represents all plots where the tree species occurred either on the open access control plot (Var C) or within the exclosure (Var E). Percentages are depicted for different points of time (years) and for three different subsamples, each covering a different window of time (i.e., subsample a: 9 years; subsample b: 15 years; subsample c: 24 years). The line *overall* represents the share of plots, where trees reached a height of 160 cm regardless of the tree species.
		Years
Tree species	Var	3			6			9			12		15		18	21	24	n
Tree species	Var	a	b	c	a	b	c	a	b	c	b	c	b	c	c	c	c	a	b	c
Abies alba	C	0.0	0.0	0.0	0.3	0.0	0.0	1.7	0.0	0.0	1.0	0.0	2.7	0.0	1.2	2.8	3.6	1130	621	241
Abies alba	E	0.0	0.0	0.0	0.1	0.0	0.0	3.6	1.5	0.8	4.0	1.2	9.7	4.0	11.6	17.9	25.1	1130	621	241
Acer platanoides	C	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	64	36	15
Acer platanoides	E	0.0	0.0	0.0	0.0	0.0	0.0	3.1	0.0	0.0	0.0	0.0	2.8	0.0	0.0	6.7	13.3	64	36	15
Acer pseudoplatanus	C	0.0	0.0	0.0	0.6	0.2	0.0	2.4	0.5	0.0	1.6	0.0	3.0	0.0	1.1	2.3	6.2	790	439	177
Acer pseudoplatanus	E	0.3	0.5	0.0	3.8	2.7	0.6	9.9	5.3	1.7	8.2	2.8	14.6	5.6	10.2	16.9	26.0	790	439	177
Betula pendula	C	0.0	0.0	0.0	1.8	0.0	0.0	7.0	0.0	0.0	1.8	0.0	1.8	0.0	0.0	9.5	9.5	114	57	21
Betula pendula	E	0.0	0.0	0.0	4.4	1.8	0.0	13.2	1.8	0.0	3.5	0.0	5.3	0.0	0.0	19.1	33.3	114	57	21
Fraxinus excelsior	C	0.3	0.6	0.0	2.2	0.6	0.0	5.3	1.2	0.0	3.5	0.0	5.8	0.0	0.0	0.0	3.2	323	173	63
Fraxinus excelsior	E	0.6	1.2	0.0	6.5	2.9	0.0	12.1	5.8	0.0	8.1	0.0	11.0	0.0	1.6	3.2	11.1	323	173	63
Fagus sylvatica	C	0.7	0.4	0.0	3.3	2.0	0.0	13.7	6.8	1.7	11.8	4.0	17.1	4.6	8.6	15.4	18.9	872	456	175
Fagus sylvatica	E	0.5	0.0	0.0	6.9	2.4	0.6	21.0	12.3	2.9	22.4	5.7	34.2	10.9	25.2	40.6	50.9	872	456	175
Picea abies	C	0.0	0.0	0.0	1.2	0.0	0.0	5.0	1.8	1.0	4.9	2.4	7.9	3.4	7.6	11.3	15.1	1216	668	291
Picea abies	E	0.3	0.0	0.0	2.4	0.8	0.0	7.0	3.2	1.0	6.3	2.1	10.8	5.8	9.3	12.7	18.2	1216	668	291
Prunus avium	C	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	5.6	0.0	0.0	0.0	0.0	43	18	5
Prunus avium	E	0.0	0.0	0.0	4.7	0.0	0.0	9.3	0.0	0.0	0.0	0.0	5.6	0.0	0.0	0.0	20.0	43	18	5
Quercus sp.	C	0.0	0.0	0.0	0.0	0.0	0.0	1.2	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	163	81	31
Quercus sp.	E	0.0	0.0	0.0	0.0	0.0	0.0	4.9	0.0	0.0	0.0	0.0	2.5	0.0	0.0	0.0	3.2	163	81	31
Sorbus aria	C	0.0	0.0	0.0	0.0	0.0	0.0	0.9	0.0	0.0	0.0	0.0	1.3	0.0	0.0	8.3	11.1	115	76	36
Sorbus aria	E	0.0	0.0	0.0	0.0	0.0	0.0	1.7	0.0	0.0	1.3	0.0	4.0	0.0	0.0	5.6	16.7	115	76	36
Sorbus aucuparia	C	0.1	0.0	0.0	0.8	0.3	0.0	1.7	0.3	0.0	0.3	0.0	0.7	0.0	0.5	0.5	2.1	728	407	191
Sorbus aucuparia	E	0.3	0.3	0.0	3.7	1.7	1.1	11.0	4.4	1.1	8.8	3.7	13.3	5.8	11.5	23.0	33.5	728	407	191
Salix caprea	C	0.0	0.0	0.0	0.0	0.0	0.0	1.4	0.0	0.0	0.0	0.0	0.0	0.0	14.3	14.3	14.3	70	29	7
Salix caprea	E	1.4	0.0	0.0	1.4	0.0	0.0	7.1	0.0	0.0	10.3	14.3	20.7	14.3	14.3	14.3	14.3	70	29	7
Salix sp.	C	0.0	0.0	0.0	2.7	0.0	0.0	2.7	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	12.5	37	17	8
Salix sp.	E	0.0	0.0	0.0	2.7	0.0	0.0	8.1	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	25.0	37	17	8
Taxus baccata	C	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	61	35	18
Taxus baccata	E	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	2.9	0.0	0.0	0.0	0.0	61	35	18
Ulmus glabra	C	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	3.9	0.0	0.0	0.0	5.9	100	52	17
Ulmus glabra	E	0.0	0.0	0.0	2.0	0.0	0.0	12.0	0.0	0.0	1.9	0.0	5.8	0.0	0.0	11.8	17.6	100	52	17
Overall	C	0.1	0.1	0.0	1.2	0.4	0.0	4.5	1.5	0.5	3.4	1.1	5.7	1.4	3.4	6.0	8.6	1722	907	332
Overall	E	0.6	0.4	0.0	7.5	2.4	0.4	24.1	6.7	1.4	14.7	5.5	28.0	8.6	15.5	31.9	57.3	1722	907	332

Tree species diversity

Ungulates might impact the occurrence and abundance of plant species by feeding on available and preferred plants, thereby shaping plant communities and biodiversity at several scales. Vice versa, plant species exert specific strategies in responses to ungulate herbivory by adapting resistance and tolerance (Rhodes et al. 2017). Studies on effects of ungulate herbivory on tree species diversity provide largely divergent results: In some cases, ungulate herbivory resulted in higher tree species diversity (Nishizawa et al. 2016) whereas in other cases it led to a reduction (Bernes et al. 2018; Reimoser et al. 2022) or had no significant impact (Nopp-Mayr et al. 2023; Royo and Carson 2022). The latter case also applied to our 30-year data set, where no significant difference in diversity measures between open access plots and exclosures could be seen. There are many reasons for this huge bandwidth of effects. For example, different ungulate species show different feeding preferences (Ingram et al. 2013), which can tip the scales in interspecies competition between tree species (McGarvey et al. 2013; Weisberg and Bugmann 2003). This is particularly crucial for tree species which are very vulnerable to browsing and do not survive intense herbivory (Nishizawa et al. 2016). Highly preferred plant species might detract herbivores from other plant species, which in turn escape feeding pressure. Furthermore, rare plant species can be highly attractive for ungulates (Bödeker et al. 2021). If, however, a plethora of alternative feeding sources are available, vulnerable tree species might not be affected by herbivory (Kupferschmid et al. 2020). Different capabilities of plant species to produce toxins, thorns, and other herbivory deterrents plays a major role in the feeding decisions of ungulates. These capabilities, which sometimes are only produced if subjected to browsing by certain herbivores, are also influenced by environmental factors (Seager et al. 2013) or a plant´s life stage (Selaković et al. 2017). Furthermore, factors like the spatial and temporal distribution of ungulates, the applied silvicultural system, stem density of young trees and more also affect herbivore impacts on woody plants (Reimoser 2003). For example, Reimoser et al. (2022) found higher overall biodiversity in a large area in the case of an uneven spatial distribution of ungulates compared to evenly distributed or absent ungulates as high spatial variability in ungulate presence also leads to variability of related browsing exposure of plants.

The complexity of these driving factors including ungulate species composition or abundance, differences in environmental conditions (e.g., soil, aspect, or altitude), differences in plant communities together with differing time periods covered by surveys explains the variety of reported impacts of ungulate herbivory on tree species diversity. However, in contrast to the lack of effects within the pooled data set in our study (including all open access and exclosure plots at larger spatial scales), there might be distinct differences in tree species diversity between plot pairs on a local scale. Therefore, additional assessments for smaller areas with similar ecological conditions are recommended if management consequences are to be drawn from the effects of ungulates on tree species diversity.

Ungulates can impact forest vegetation by trampling (including pawing and scraping), fraying, bark-peeling and, perhaps most importantly, browsing (both visible browsing and unseen browsing, such as feeding on seeds and seedlings). Intense selective browsing of certain tree species often results in reduced regeneration of these species, thereby changing the overall tree species composition. However, such impacts do not automatically mean damage to the forest. In forestry, the occurrence of damage or benefit depends on the respective objective for the forest. Damage in forestry terms may arise if the ungulate impact leads to a reduction in forest growth, economic value, ecological stability, species diversity, or in a forest’s protective function (e.g. against avalanches, land-slides, erosion, floods) (Reimoser et al. 2023). Beneficial effects can result from increasing diversity and improved economic value by ungulates selectively browsing non-target plant species and by improving germination conditions (seed dispersal, impact of hooves, and droppings; cf. Gill and Beardall 2001; Howe and Westley 1988; Putman 1986, 1996; Reimoser 2003).

Height class diversity of trees

The analysis of height class diversity as a proxy for vertical structural diversity of trees on the survey plots resulted in lower Shannon and Simpson diversity on control plots than on exclosure plots, indicating that ungulate herbivory reduced the vertical structural diversity. This result was in line with an earlier study from another Austrian alpine region (Nopp-Mayr et al. 2023) and might be partially due to the fact, that continuous browsing of the terminal shoot over long periods of time hampers height growth (Beguin et al. 2016). Therefore, most trees on control plots were attributed to a small number of different height classes in the given study. On exclosure plots, other factors than herbivory, including inter- and intraspecies competition combined with environmental factors might determine height growth patterns, which leads to a differentiation in vertical structure. Again, the multitude of co-occurring processes that shape height structures of plant communities including preferences or avoidance of specific plant species by ungulates (Burney and Jacobs 2011; Provenza et al. 1992, 2003), the duration and intensity of browsing, the seasonally changing food availability, or ungulates' preferences for certain plant heights (Hansen and Clark 1977; Nelson and Leege 1982) might induce different modifications of structural features of plant communities by ungulate feeding. Additionally, plants try to vertically escape ungulate herbivory by height increment, by changing leaf traits (photosynthesis, morphology, and chemistry) at specific height strata (Rhodes et al. 2017; Sabo 2019) or plant architecture (Johnston et al. 2007), yielding non-uniform structural responses of plant communities to ungulate’s feeding. It should be mentioned that height classes in our study did not cover constant ranges, as smaller ranges were used for the first life stages of trees. This means that, while it is possible to draw conclusions on vertical structural diversity with height class diversity, it is not possible to infer on absolute height differences. With regard to achieving forestry goals in forest management, the differences observed in the diversity of vertical structure are unlikely to play a major role.

Height class distribution of trees

As differences in absolute height growth, caused by ungulate herbivory, are potentially important for forest regeneration and later developmental phases, the heights of the six highest tree individuals per tree species and per plot were analyzed. As was to be expected, trees on exclosure plots were significantly higher compared to open control plots after a survey time of 30 years. The reason for this development over the 30 years is the repeated herbivory of ungulates, which reduces height growth, especially if terminal shoots are repeatedly browsed (Hood and Bayley 2009; Kupferschmid et al. 2019). This difference in height growth could have a particularly negative impact in our study area, as almost half of the state's forests are protective forests against avalanches and rockfalls, where fast forest regeneration is of key importance. In such steep protective forests above settlements and roads, ungulates are hunted very intensively, if necessary, all year round.

While browsing of terminal shoots in a single year does not necessarily induce height growth suppression in individual trees, but might even promote height growth, longer lasting herbivory frequently results in a height growth delay of browsed trees (e.g., Kupferschmid et al. 2022; Siipilehto and Heikkilä 2005), lowering their competitive power (Kupferschmid et al. 2014). In our study, browsing of ungulates did not prevent forest trees from growing with time, but we observed a significantly differing height growth performance at year 30 of the trial.

Shrub species diversity

In our study, overall shrub species diversity did not differ significantly between control and exclosure plots after a survey span of 30 years, which is in line with results from another long-term survey in the Austrian alps (Nopp-Mayr et al. 2023). However, earlier studies on the effects of ungulate herbivory on shrub species diversity yielded very different outcomes (see Bernes et al. 2018). Compared to other plant strata and taxa, deciduous woody plants show distinct seasonal changes in the amount of available biomass for large herbivore feeding, simultaneously being a crucial food resource during periods with snow cover (Cook 2002; Jenkins and Wright 1987). However, co-occurring woody plant species potentially compete for resources and intensive herbivore feeding on one species or specific individuals might shape inter- and intraspecific competition for the others (Krueger et al. 2009). Although herbivore´s preferences depend on the specific ecosystem context and the foraging types of ungulate species (see Shipley et al. 2009), they can be expected to have different effects of herbivory on the diversity of shrub species compared to tree species, but this was not the case in our study.

Reaching the critical tree height - open access vs. exclosures

Species-specific and context-dependent growth patterns of tree species (Rhodes et al. 2017) determine the time span, where terminal shoots of plants are accessible for ungulate herbivores as well as the chance for outgrow. The time it takes for trees to outgrow ungulate herbivory ties back into the afore-discussed height growth differences. Due to the slower growth on control plots, it takes trees longer to reach the 160 cm threshold with their terminal shoots than on exclosure plots. This is clearly expressed by a lower percentage of control plots where trees have grown out of most vulnerable height classes at a given time in the vast majority of cases.

Potential bias of long-term survey in terms of height growth performance

The reduction in height growth caused by ungulates was further cemented when comparing data sets with different survey duration. The longer our survey was running, the higher was the proportion of sites with lower height growth performance although the survey concept included trees being higher than the usual feeding height of wild ungulates. Investigating all plot pairs from 9 years on resulted in earlier gains of plots with trees overgrowing the 160 cm threshold, compared to only observing the 24- year old plot pairs, which showed very late outgrowth. This was the case both for control and exclosure plots. In other words, height growth on long-standing plot pairs was slower than on plot pairs that did not stand for such long periods. Thus, long-term control-exclosure data sets indeed inherently carry a potential bias towards specific site conditions. This includes several factors, that drive the feasibility of long-term surveys and the number of plot pairs persisting over a time period of 30 years: (1) In mountainous terrain, natural disturbances like wind throws, avalanches and rockfall as well as material aging have a strong influence on the functional life of exclosures. This might strongly correlate with slope and other terrain features. Once the effectiveness of exclosures is impaired and ungulates gain access, the related plot pair must be abandoned. (2) Long-term comparisons between exclosures and open access plots require no or at least equal forest management in the immediate surroundings of individual plot pairs, which might not easily be ensured. In case that forest operations in the adjacent mature stand impact on of the two related plots in a different way, the plot pair has to be omitted. This also applies to natural abiotic disturbances. (3) Long-term surveys often cannot be carried out in a scientific or university context alone, but require the involvement of local stakeholders. This entails the risk that surveys are terminated once regeneration reaches a certain tree height. The current pressure of ungulates on forest regeneration is a common calculation basis for annual game harvest plans, which was also the case in our study. Therefore, plots that had outgrown 160 cm were of less interest for some practitioners within an ungulate impact monitoring scheme and some plot pairs had thus no longer been recorded or abolished. It must be said, however, that old slow-growing sites also show a skewed current impact of ungulates as they have been impacted by ungulate herbivory for many years prior and therefore are also a less reliable calculation basis for the new annual game harvest plans. It can be assumed, that all these factors have an influence on our data set although the magnitude of their impact cannot be easily delineated.

The effects of ungulates on biodiversity of forest vegetation vary depending on the diversity indicators used. Basically, forest structure and dynamics respond to the density of ungulates as well as many other influencing factors in the forest environment. In order to filter out the lasting effects of ungulates from the entire complex of interacting impacts, long-term studies with ungulate exclusions are urgently required. Although the exclosures do not represent a natural situation by completely preventing access for ungulates, they are methodologically indispensable.

However, long-term studies over several decades have the disadvantage that the ungulate situation at the beginning of the monitoring can be completely different from that at the end of the study period, depending on the management measures taken. This makes the interpretation of the results more difficult. Additionally, there is a risk that at the end of the study, not all pairs of plots comparable to the initial situation will be available. If the plot-pairs that have been lost over the decades are not randomly distributed, the final results for the entire area to be studied are not directly comparable, as the selective loss of certain sample areas can lead to a bias of the results.

If long-term surveys are not carried out by scientists but by stakeholders, it must be ensured that the surveys are continued even if the time periods relevant to some specific management issues have already been exceeded. This is a crucial and challenging point in a long-term survey setting.

Short-term studies on the influence of ungulates on forest development have the advantage that they always refer to current ungulate populations, making them easier to interpret, and an efficient adaptive management is possible. Studies with only short-term monitoring results are also easier to finance. However, the long-term overall impact of ungulates on forest development with the complex compensation and reinforcement effects over time, remain unknown. If comparing studies with different durations, one should be aware of these pros and cons.

Author Contribution

S.R. and F.R. designed and conducted the field work and provided all data. U.N.-M., W.L., S.R. and F.R. conceptualized the study. U.N.-M. and W.L. selected and conducted the statistical analyses, assisted by S.R. and F.R.. U.N.-M., W.L. and F.R. wrote the main manuscript text. All authors reviewed the manuscript.

Conflict of interest statement

The authors declare no conflict of interest.

Acknowledgements

We thank all the persons in the state authority, in the forestry companies, and the land owners who made this long-term study possible. We are also grateful to Dr. Eva Maria Schöll MSc for statistical advice.

Abbas F, Merlet J, Morellet N, et al (2012) Roe deer may markedly alter forest nitrogen and phosphorus budgets across Europe. Oikos 121:1271–1278. https://doi.org/10.1111/j.1600-0706.2011.20103.x
Adams TP, W. Purves D, W. Pacala S (2007) Understanding height-structured competition in forests: is there an R* for light? Proceedings of the Royal Society B: Biological Sciences 274:3039–3048. https://doi.org/10.1098/rspb.2007.0891
Aguirrebengoa M, García-Planas M, Müller C, González-Megías A (2018) Transgenerational effects of ungulates and pre-dispersal seed predators on offspring success and resistance to herbivory. PLoS ONE 13:e0207553. https://doi.org/10.1371/journal.pone.0207553
Beguin J, Tremblay J-P, Thiffault N, et al (2016) Management of forest regeneration in boreal and temperate deer–forest systems: challenges, guidelines, and research gaps. Ecosphere 7:e01488. https://doi.org/10.1002/ecs2.1488
Bellingham PJ, Allan CN (2003) Forest regeneration and the influences of white-tailed deer (Odocoileus virginianus) in cool temperate New Zealand rain forests. Forest Ecology and Management 175:71–86
Bernes C, Macura B, Jonsson BG, et al (2018) Manipulating ungulate herbivory in temperate and boreal forests: effects on vegetation and invertebrates. A systematic review. Environ Evid 7:13. https://doi.org/10.1186/s13750-018-0125-3
Bödeker K, Ammer C, Knoke T, Heurich M (2021) Determining Statistically Robust Changes in Ungulate Browsing Pressure as a Basis for Adaptive Wildlife Management. Forests 12:1030. https://doi.org/10.3390/f12081030
Burney OT, Jacobs DF (2011) Ungulate herbivory of regenerating conifers in relation to foliar nutrition and terpenoid production. Forest Ecology and Management 262:1834–1845. https://doi.org/10.1016/j.foreco.2011.07.035
Casabon C, Pothier D (2007) Browsing of tree regeneration by white-tailed deer in large clearcuts on Anticosti Island, Quebec. Forest Ecology and Management 253:112–119. https://doi.org/10.1016/j.foreco.2007.07.009
Chao A, Gotelli NJ, Hsieh TC, et al (2014) Rarefaction and extrapolation with Hill numbers: a framework for sampling and estimation in species diversity studies. Ecological Monographs 84:45–67. https://doi.org/10.1890/13-0133.1
Chao A, Kubota Y, Zelený D, et al (2020) Quantifying sample completeness and comparing diversities among assemblages. Ecological Research 35:292–314. https://doi.org/10.1111/1440-1703.12102
Cilles SE, Coy G, Stieha CR, et al (2016) A Comparison of Seed Predation, Seed Dispersal, and Seedling Herbivory in Oak and Hickory: Species with Contrasting Regenerating Abilities in a Bluegrass Savanna–Woodland Habitat. Northeastern Naturalist 23:466–481
Colwell RK, Chao A, Gotelli NJ, et al (2012) Models and estimators linking individual-based and sample-based rarefaction, extrapolation and comparison of assemblages. Journal of Plant Ecology 5:3–21. https://doi.org/10.1093/jpe/rtr044
Cook JG (2002) Nutrition and Food. In: Toweill DE, Thomas JW (eds) North American Elk: Ecology and Management. Smithsonian Institution Press, Washington, DC, USA, pp 259–349
Gaquerel E, Gulati J, Baldwin IT (2014) Revealing insect herbivory-induced phenolamide metabolism: from single genes to metabolic network plasticity analysis. The Plant Journal 79:679–692. https://doi.org/10.1111/tpj.12503
Gill RMA, Beardall V (2001) The impact of deer on woodlands: the effects of browsing and seed dispersal on vegetation structure and composition. Forestry: An International Journal of Forest Research 74:209–218. https://doi.org/10.1093/forestry/74.3.209
Hansen RM, Clark RC (1977) Foods of Elk and Other Ungulates at Low Elevations in Northwestern Colorado. The Journal of Wildlife Management 41:76–80. https://doi.org/10.2307/3800093
Hedwall P-O, Churski M, Jędrzejewska B, et al (2018) Functional composition of temperate forest trees under chronic ungulate herbivory. Journal of Vegetation Science 29:179–188. https://doi.org/10.1111/jvs.12623
Hill MO (1973) Diversity and Evenness: A Unifying Notation and Its Consequences. Ecology 54:427–432. https://doi.org/10.2307/1934352
Holeski LM, McKenzie SC, Kruger EL, et al (2016) Phytochemical traits underlie genotypic variation in susceptibility of quaking aspen (Populus tremuloides) to browsing by a keystone forest ungulate. Journal of Ecology 104:850–863. https://doi.org/10.1111/1365-2745.12559
Hood GA, Bayley SE (2009) A comparison of riparian plant community response to herbivory by beavers (Castor canadensis) and ungulates in Canada’s boreal mixed-wood forest. Forest Ecology and Management 258:1979–1989. https://doi.org/10.1016/j.foreco.2009.07.052
Howe H, Westley L (1988) Ecological relationships of plants and animals. Oxford University Press Inc., University of California
Hsieh TC, Ma KH, Chao A (2016) iNEXT: an R package for rarefaction and extrapolation of species diversity (Hill numbers). Methods in Ecology and Evolution 7:1451–1456. https://doi.org/10.1111/2041-210X.12613
Ingram RS, Doran MP, Nader G, et al (2013) Planned Herbivory in the Management of Wildfire Fuels. In: Herbivory. IntechOpen
Jenkins KJ, Wright RG (1987) Dietary niche relationships among cervids relative to winter snowpack in northwestern Montana. Can J Zool 65:1397–1401. https://doi.org/10.1139/z87-220
Johnston DB, Cooper DJ, Hobbs NT (2007) Elk browsing increases aboveground growth of water-stressed willows by modifying plant architecture. Oecologia 154:467–478. https://doi.org/10.1007/s00442-007-0854-4
Kay CE, Bartos DL (2000) Ungulate Herbivory on Utah Aspen: Assessment of Long-Term Exclosures. Journal of Range Management 53:145–153. https://doi.org/10.2307/4003274
Krueger LM, Peterson CJ, Royo A, Carson WP (2009) Evaluating relationships among tree growth rate, shade tolerance, and browse tolerance following disturbance in an eastern deciduous forest. Can J For Res 39:2460–2469. https://doi.org/10.1139/X09-155
Kupferschmid AD, Brang P, Bugmann H (2019) Abschätzung des Einflusses von Verbiss durch wildlebende Huftiere auf die Baumverjüngung. Schweizerische Zeitschrift fur Forstwesen 170:125–134. https://doi.org/10.3188/szf.2019.0125
Kupferschmid AD, Bütikofer L, Hothorn T, et al (2020) Ungulate Species and Abundance as well as Environmental Factors Determine the Probability of Terminal Shoot Browsing on Temperate Forest Trees. Forests 11:764. https://doi.org/10.3390/f11070764
Kupferschmid AD, Greilsamer R, Brang P, et al (2022) Assessment of the Impact of Ungulate Browsing on Tree Regeneration. IntechOpen
Kupferschmid AD, Wasem U, Bugmann H (2014) Light availability and ungulate browsing determine growth, height and mortality of Abies alba saplings. Forest Ecology and Management 318:359–369. https://doi.org/10.1016/j.foreco.2014.01.027
Lertzman KP (1992) Patterns of Gap-Phase Replacement in a Subalpine, Old-Growth Forest. Ecology 73:657–669. https://doi.org/10.2307/1940772
Manso R, McLean JP, Arcangeli C, Matthews R (2021) Dynamic top height models for several major forest tree species in Great Britain. Forestry: An International Journal of Forest Research 94:181–192. https://doi.org/10.1093/foresj/cpaa036
Massad TJ (2013) Ontogenetic differences of herbivory on woody and herbaceous plants: a meta-analysis demonstrating unique effects of herbivory on the young and the old, the slow and the fast. Oecologia 172:1–10. https://doi.org/10.1007/s00442-012-2470-1
Matisons R, Kārkliņa A, Krišāns O, et al (2020) Species composition modulates seedling competitiveness of temperate tree species under hemiboreal conditions. Forest Ecology and Management 478:118499. https://doi.org/10.1016/j.foreco.2020.118499
Maxwell JD, Rhodes AC, St. Clair SB (2019) Human altered disturbance patterns and forest succession: impacts of competition and ungulate herbivory. Oecologia 189:1061–1070. https://doi.org/10.1007/s00442-019-04370-8
McGarvey JC, Bourg NA, Thompson JR, et al (2013) Effects of Twenty Years of Deer Exclusion on Woody Vegetation at Three Life-History Stages in a Mid-Atlantic Temperate Deciduous Forest. Northeastern Naturalist 20:451–468
Millett J, Edmondson S (2013) The impact of 36 years of grazing management on vegetation dynamics in dune slacks. Journal of Applied Ecology 50:1367–1376. https://doi.org/10.1111/1365-2664.12113
Murray BD, Webster CR, Bump JK (2013) Broadening the ecological context of ungulate–ecosystem interactions: the importance of space, seasonality, and nitrogen. Ecology 94:1317–1326. https://doi.org/10.1890/12-1582.1
Nelson J, Leege T (1982) Nutritional Requirements and Food Habits. Elk of North America: Ecology and Management 323–367
Nishizawa K, Tatsumi S, Kitagawa R, Mori AS (2016) Deer herbivory affects the functional diversity of forest floor plants via changes in competition-mediated assembly rules. Ecological Research 31:569–578. https://doi.org/10.1007/s11284-016-1367-6
Nopp-Mayr U, Reimoser S, Reimoser F, et al (2020) Analyzing long-term impacts of ungulate herbivory on forest-recruitment dynamics at community and species level contrasting tree densities versus maximum heights. Sci Rep 10:20274. https://doi.org/10.1038/s41598-020-76843-3
Nopp-Mayr U, Schöll EM, Sachser F, et al (2023) Does Ungulate Herbivory Translate into Diversity of Woody Plants? A Long-Term Study in a Montane Forest Ecosystem in Austria. Diversity 15:165. https://doi.org/10.3390/d15020165
Payton ME, Greenstone MH, Schenker N (2003) Overlapping confidence intervals or standard error intervals: What do they mean in terms of statistical significance? insc 3:1–6. https://doi.org/10.1673/031.003.3401
Pellerin M, Saïd S, Richard E, et al (2010) Impact of deer on temperate forest vegetation and woody debris as protection of forest regeneration against browsing. Forest Ecology and Management 260:429–437. https://doi.org/10.1016/j.foreco.2010.04.031
Perkovich C, Ward D (2022) Differentiated plant defense strategies: Herbivore community dynamics affect plant–herbivore interactions. Ecosphere 13:e3935. https://doi.org/10.1002/ecs2.3935
Persson I-L, Danell K, Bergström R (2005) Different Moose Densities and Accompanied Changes in Tree Morphology and Browse Production. Ecological Applications 15:1296–1305. https://doi.org/10.1890/04-0499
Provenza F, Pfister J, Cheney C (1992) Mechanisms of Learning in Diet Selection with Reference to Phytotoxicosis in Herbivores. Journal of Range Management 36:36–45. https://doi.org/10.2307/4002523
Provenza FD, Villalba JJ, Dziba LE, et al (2003) Linking herbivore experience, varied diets, and plant biochemical diversity. Small Ruminant Research 49:257–274. https://doi.org/10.1016/S0921-4488(03)00143-3
Putman R (1986) Grazing in Temperate Ecosystems: Large Herbivores and the Ecology of the New Forest. Springer Netherlands
Putman RJ (1996) Ungulates in temperate forest ecosystems: perspectives and recommendations for future research. Forest Ecology and Management 88:205–214. https://doi.org/10.1016/S0378-1127(96)03878-9
Reimoser F (2003) Steering the impacts of ungulates on temperate forests. Journal for Nature Conservation 10:243–252. https://doi.org/10.1078/1617-1381-00024
Reimoser F, Armstrong H, Suchant R (1999) Measuring forest damage of ungulates: what should be considered. Forest Ecology and Management 120:47–58. https://doi.org/10.1016/S0378-1127(98)00542-8
Reimoser F, Reimoser S, Zsak K (2022) Long-term impact of wild ungulates on natural forest regeneration in the Donau-Auen National Park, Austria. 158:97–127
Reimoser S, Nopp-Mayr U, Reimoser F (2023) Ungulate exclosures (UICOS method) as a monitoring tool for integrative forest-ungulate management in Central Europe – method description
Rhodes AC, Anderson V, St Clair SB (2017) Ungulate herbivory alters leaf functional traits and recruitment of regenerating aspen. Tree Physiology 37:402–413. https://doi.org/10.1093/treephys/tpx015
Royo AA, Carson WP (2022) Stasis in forest regeneration following deer exclusion and understory gap creation: A 10-year experiment. Ecological Applications 32:e2569. https://doi.org/10.1002/eap.2569
Russell FL, Fowler NL (2004) Effects of white-tailed deer on the population dynamics of acorns, seedlings and small saplings of Quercus buckleyi. Plant Ecology 173:59–72
Sabo AE (2019) Impacts of Browsing and Grazing Ungulates on Plant Characteristics and Dynamics. In: Gordon IJ, Prins HHT (eds) The Ecology of Browsing and Grazing II. Springer International Publishing, Cham, pp 259–276
Schowalter TD (2016) Chapter 12 - Herbivory. In: Schowalter TD (ed) Insect Ecology (Fourth Edition). Academic Press, pp 405–443
Schrijvers-Gonlag M, Skarpe C, Andreassen HP (2020) Influence of light availability and soil productivity on insect herbivory on bilberry (Vaccinium myrtillus L.) leaves following mammalian herbivory. PLoS ONE 15:e0230509. https://doi.org/10.1371/journal.pone.0230509
Seager ST, Eisenberg C, St. Clair SB (2013) Patterns and consequences of ungulate herbivory on aspen in western North America. Forest Ecology and Management 299:81–90. https://doi.org/10.1016/j.foreco.2013.02.017
Sebata A (2013) Woody Plant-Herbivore Interactions in Semi-Arid Savanna Ecosystems. In: Herbivory. IntechOpen
Selaković S, Vujić V, Stanisavljević N, et al (2017) Ontogenetic stage, plant vigor and sex mediate herbivory loads in a dioecious understory herb. Acta Oecologica 85:184–190. https://doi.org/10.1016/j.actao.2017.11.006
Seymour CL, Joseph GS, Makumbe M, et al (2016) Woody species composition in an African savanna: determined by centuries of termite activity but modulated by 50 years of ungulate herbivory. Journal of Vegetation Science 27:824–833. https://doi.org/10.1111/jvs.12393
Shipley LA, Forbey JS, Moore BD (2009) Revisiting the dietary niche: When is a mammalian herbivore a specialist? Integrative and Comparative Biology 49:274–290. https://doi.org/10.1093/icb/icp051
Siipilehto J, Heikkilä R (2005) The effect of moose browsing on the height structure of Scots pine saplings in a mixed stand. Forest Ecology and Management 205:117–126. https://doi.org/10.1016/j.foreco.2004.10.051
Tschöpe O, Wallschläger D, Burkart M, Tielbörger K (2011) Managing open habitats by wild ungulate browsing and grazing: A case-study in North-Eastern Germany: Managing open habitats by wild ungulate browsing and grazing. Applied Vegetation Science 14:200–209. https://doi.org/10.1111/j.1654-109X.2010.01119.x
VanderMolen MS, Webster CR (2021) Influence of deer herbivory on regeneration dynamics and gap capture in experimental gaps, 18 years post-harvest. Forest Ecology and Management 501:119675. https://doi.org/10.1016/j.foreco.2021.119675
Walters MB, Farinosi EJ, Willis JL (2020) Deer browsing and shrub competition set sapling recruitment height and interact with light to shape recruitment niches for temperate forest tree species. Forest Ecology and Management 467:118134. https://doi.org/10.1016/j.foreco.2020.118134
Weisberg PJ, Bugmann H (2003) Forest dynamics and ungulate herbivory: from leaf to landscape. Forest ecology and management 181:1–12
Wilkinson L (2011) ggplot2: Elegant Graphics for Data Analysis by WICKHAM, H. Biometrics 67:678–679. https://doi.org/10.1111/j.1541-0420.2011.01616.x
Williams PJ, Brodie JF (2023) Predicting how defaunation-induced changes in seed predation and dispersal will affect tropical tree populations. Conservation Biology 37:e14014. https://doi.org/10.1111/cobi.14014
Wisdom MJ, Vavra M, Boyd JM, et al (2006) Understanding Ungulate Herbivory—Episodic Disturbance Effects on Vegetation Dynamics: Knowledge Gaps and Management Needs. Wildlife Society Bulletin 34:283–292. https://doi.org/10.2193/0091-7648(2006)34[283:UUHDEO]2.0.CO;2
Zamora R, Gómez JM, Hódar JA, et al (2001) Effect of browsing by ungulates on sapling growth of Scots pine in a Mediterranean environment: consequences for forest regeneration. Forest Ecology and Management 144:33–42
Zhou S, Lou Y-R, Tzin V, Jander G (2015) Alteration of plant primary metabolism in response to insect herbivory. Plant Physiol pp.01405.2015. https://doi.org/10.1104/pp.15.01405

No competing interests reported.

ESM1.pdf

Download PDF

Reviewers agreed at journal
19 Aug, 2024
Reviewers invited by journal
14 Aug, 2024
Editor assigned by journal
06 Aug, 2024
Submission checks completed at journal
02 Aug, 2024
First submitted to journal
01 Aug, 2024

You are reading this latest preprint version

Long-term surveys of ungulates’ effects on tree and shrub species in mountainous forests – a valuable outcome or a biased picture?

Status:

Version 1

Abstract

Figures

Introduction

Materials and Methods

Study area

Field Sampling

Data preparation

Statistical analyses

Results

Tree species diversity

Height class diversity of trees

Height class distribution

Shrub species diversity

Temporal pattern of the proportion of plots where terminal shoots outgrow the ungulate browsing height

Control plots vs. exclosures

Short-term vs. long-term survey duration

Discussion

Tree species diversity

Height class diversity of trees

Height class distribution of trees

Shrub species diversity

Reaching the critical tree height - open access vs. exclosures

Potential bias of long-term survey in terms of height growth performance

Conclusion

Declarations

Author Contribution

References

Additional Declarations

Supplementary Files

Status:

Version 1