Climate and air pollution impacts on habitat suitability of Austrian forest ecosystems
Climate and air pollution impacts on habitat suitability of Austrian forest ecosystems
Thomas DirnboÈ ck 0 1
Ika Djukic 0 1
Barbara Kitzler 1
Johannes Kobler 0 1
Janet P. Mol- Dijkstra 1
Max Posch 1
Gert Jan Reinds 1
Angela Schlutow 1
Franz Starlinger 1
Wieger G. W. Wamelink 1
0 Department for Ecosystem Research and Environmental Information Management, Environment Agency Austria , Spittelauer LaÈnde 5, Vienna , Austria , 2 Austrian Research Centre for ForestsÐBFW , Vienna, Austria, 3 Alterra, Wageningen UR , Wageningen, the Netherlands, 4 Coordination Centre for Effects (CCE), RIVM, Bilthoven, the Netherlands, 5 OEKO-DATAÐEcosystem Analysis and Environmental Data Management , Strausberg , Germany
1 Editor: RunGuo Zang, Chinese Academy of Forestry , CHINA
Climate change and excess deposition of airborne nitrogen (N) are among the main stressors to floristic biodiversity. One particular concern is the deterioration of valuable habitats such as those protected under the European Habitat Directive. In future, climate-driven shifts (and losses) in the species potential distribution, but also N driven nutrient enrichment may threaten these habitats. We applied a dynamic geochemical soil model (VSD+) together with a novel niche-based plant response model (PROPS) to 5 forest habitat types (18 forest sites) protected under the EU Directive in Austria. We assessed how future climate change and N deposition might affect habitat suitability, defined as the capacity of a site to host its typical plant species. Our evaluation indicates that climate change will be the main driver of a decrease in habitat suitability in the future in Austria. The expected climate change will increase the occurrence of thermophilic plant species while decreasing cold-tolerant species. In addition to these direct impacts, climate change scenarios caused an increase of the occurrence probability of oligotrophic species due to a higher N immobilisation in woody biomass leading to soil N depletion. As a consequence, climate change did offset eutrophication from N deposition, even when no further reduction in N emissions was assumed. Our results show that climate change may have positive side-effects in forest habitats when multiple drivers of change are considered.
Data Availability Statement: All relevant data are
included within the paper and its Supporting
Funding: The study was funded by the Austrian
Climate Research Program (ACRP5 - CCN-Adapt
KR12AC5K01308) and by the Austrian Federal
Ministry of Agriculture, Forestry. We are grateful to
the data providers of the UNECE-CLRTAP
International Cooperative Programmes ICP Forests
and ICP Integrated Monitoring.
Climate change combined with excess deposition of nitrogen (N) are among the main stressors
of biodiversity, at least in Europe, North America and parts of Asia [
]. Climate warming has
caused phenological, physiological and genetic adaptations and changes of spatial distribution
patterns of plant and animal species [
]. Nutrient enrichment in response to N deposition
has also caused changes in the structure of communities and declines in biodiversity .
According to large-scale modelling, major parts of the European Natura 2000 habitats, which
are at the core of the European Habitat Directive [
], are under threat [
]. If both greenhouse
gas emissions, causing climatic warming, and N emissions from fossil fuel burning and
agriculture, leading to the deposition of reactive N, are not mitigated more effectively in the future,
even harsher effects are to be expected [7±9]. Some success in reducing N emissions has
already been achieved in North America and Europe. As a result, total N in wet deposition in
Europe decreased on average by 2.7% between 2001±2002 and 2005±2007 [
]. However, no
climate change mitigation measures are in place at a scale necessary to avoid further damage
] nor will N emission reductions be able to fully release sensitive habitats from chronic N
]. Moreover, N deposition in South America, Africa and most prominently in Asia are
expected to raise further [
]. That is why studying the magnitude of these threats to protected
habitats is important.
To date, assessments of potential future risks to biodiversity have largely neglected
interactions between climate change, N deposition, and biodiversity [
]. Since most plant species
have a distinct climate niche, climate warming and changes in the precipitation regime alter
the distribution pattern of their suitable habitats and eventually lead to the loss of species [13±
15]. The main effect due to increased N deposition in terrestrial ecosystems is enhanced
growth of some species, effectively using the additional N [
]. In forests, increasing
availability of soil N for plants might lead to more homogenous forest floor plant communities and
hence to biodiversity loss  which in turn might have consequences for the food web
structure and ecosystem functioning [
]. Beside deposition of inorganic N compounds, plant N
availability is controlled by the effects of temperature and moisture on litter decomposition
rates and mineralization of organic N soils [
]. Whereas N deposition leads to an increase in
N availability [20±24], expected climate change may have different effects. Warming will lead
to enhanced N mineralization and nitrification and eventually to higher N uptake in response
to higher tree growth [
]. McDonnell et al.  showed in a modelling study that tree
growth stimulation by climate change can considerably offset negative effects of N deposition
on biodiversity. Wamelink et al.  showed that a higher N deposition may lead to
devastating effect on biodiversity, especially in heathland and semi-natural grasslands.
Although we focus on eutrophication effects of N deposition and climate change induced
impacts, soil acidification (through N and sulphur deposition) is an important additional
driver. Sulphur (S) deposition (and to some extent N) has led to large-scale soil acidification
during the second half of the 20th century, but significantly decreased in Europe since the late
] followed by increasing soil pH values [
] and a decrease in the prevalence of
plant species adapted to acid soils [31±33]. Climate warming can potentially accelerate soil
recovery from acidification, because base cation input to the soil increases with an increase in
weathering and litter decomposition [
Coupled soil-plant models capable of quantifying potential interacting effects of climate
change and N deposition on biodiversity are just emerging [
] and are being increasingly
applied [27, 35, 37±39]. Here we used the dynamic biochemical soil model VSD+ [
with two plant response models, PROPS [
] and BERN (for evaluation only) [
] with data
from 18 long-term forest ecosystem observation sites distributed across Austria. Both plant
models use empirical environmental response functions to predict species occurrence based
on its realized niche [
]. We focused on habitat suitability, i.e. the capacity of a site (habitat)
to host its characteristic plant species relying on the concept of potential natural vegetation
]. We do so because the European Habitat Directive [
] adheres to the concept of
ªfavourable conservation statusº which can be measured by habitat suitability [
]. For future
predictions three different N and S deposition and four different climate scenarios were used.
Specifically, we hypothesized 1) that climate change will weaken the suitability of forest
habitats to host their characteristic plant species because major changes in temperature and
2 / 16
precipitation will push these species out of their optimum range, 2) that N deposition will also
lower the habitat suitability due to soil N enrichment and a concomitant increase in the
occurrence probability of nitrophilic plant species on the expense of oligotrophic species, and 3) that
when taking interactions into account, eutrophication in forest floor species composition due
to N deposition effects will be mitigated by climate driven increased N immobilisation in
The field sites are all part of the Austrian federal long-term monitoring program of air
pollution effects of the Austrian Ministry for Agriculture, Forestry, Environment and Water
Management. Owners of the land gave permission to conduct monitoring. Our study did not
involve field studies on endangered species other than observation no involvement of human
participants, specimens or tissue samples, or vertebrate animals, embryos or tissues.
We used 18 forest sites which are part of ICP Forests and ICP Integrated Monitoring program
within the framework of effects monitoring of the UNECE Convention on Long-range
Transboundary Air Pollution [
] (Table 1). These sites were located to represent the variation in
the more widespread forest ecosystems in Austria and hold long-term measurements of
climate parameters, stand properties and soil condition [46±48].
The sites comprise 5 forest types protected under the European Habitat Directive in
Austria. These are:
· Luzulo-Fagetum beech forests (EU code 9110) with Fagus sylvatica L. as the dominant tree
species, found at submontane to montane altitudes on acid soils in the Austrian Alps, its
foothills, and the Bohemian Massive;
· Asperulo-Fagetum beech forests (EU code 9130) with Fagus sylvatica as the dominant tree
species found on neutral or near-neutral soils with mull humus;
· Medio-European limestone beech forests of the Cephalanthero-Fag ion (EU code 9150) with
Fagus sylvatica as the dominant tree species found on dry south slopes at submontane to
lower montane altitudes with base-rich soils on carbonate bedrock;
· Pannonic woods with Quercus petraea (Matt.) Liebl. and Carpinus betulus L. (EU code 91G0),
which are mixed deciduous forests found at colline to submontane altitudes of eastern
Austria with dry continental climate;
· Acidophilous Picea forests of the montane to alpine levels (Vaccinio-Piceetea) (EU code
9410) with Picea abies (L.) H.Karst. as the dominant tree species found at montane and
subalpine altitudes of the Austrian Alps.
We use the dynamic geochemical soil model VSD+ (version 5.4, [
]) together with a newly
developed plant response models PROPS [
]. For validation purpose, we also applied the
plant model BERN [
] (Fig 1). The VSD+ model includes cation exchange (Gaines-Thomas
or Gapon) and organic C and N dynamics according to the RothC-Model (version 26.3, [
VSD+ is driven by time series of N and S deposition as well as temperature and hydrology to
predict soil solution chemistry and C and N pools. In its current version, the PROPS model is
3 / 16
EU Habitat type
9110: Luzulo-Fagetum beech forests
9130: Asperulo-Fagetum beech forests
9150: Medio-European limestone beech
forests of the Cephalanthero-Fagion
91G0: Pannonic woods with Quercus
petraea and Carpinus betulus
9410: Acidophilous Picea forests of the
montane to alpine levels
Fig 1. Model chain. Model chain with VSD+ as the dynamic soil chemistry model and PROPS model for
predicting the probability of occurrence for all characteristic species (BERN model for validation) and its
respective Habitat Suitability Index (HSI). Climate driven inputs to VSD+ came from the hydrological model
MetHyd, litter element input and plant uptake were calculated with GrowUp. Climate, N and S deposition
scenario data was input to the model chain.
4 / 16
a database holding statistical niche functions for 4053 plant species occurring in Europe that
were derived from a huge set of vegetation releveÂs together with associated soil data [
outputs of PROPS are probabilities of species occurrences as a function of precipitation,
temperature, N deposition, soil C:N ratio and soil pH. As the final habitat response indicator, we
applied the Habitat Suitability Index (HSI) that describes the degree of suitability of site
conditions for a set of typical species to co-occur. The HSI is defined as the arithmetic mean of the
normalised probabilities of occurrence of these species [
]. In our study, we adopted the
common approach taken in the EU [
] to use phytosociological plant community descriptions to
define distinctive plant species for each of the 18 forest ecosystems (see Table 1 and S1 Table).
These distinctive species are the characteristic species and all constant attendant species that
can be found with a similar abundance in more than 70% of all vegetation releveÂs representing
the plant community.
We had climate measurements with a daily resolution at the sites (6 sites) or from the closest
meteorological station representative for the site (12 sites). These time series comprised the
last 20±27 years. By using the weather generator ClimGen [
] and this measured data
(precipitation, maximum temperature, minimum temperature, air humidity, global solar radiation),
we derived the baseline climate time series for 1950±2100. Then, scenario data were
synthesized by means of anomalies gathered from the A1B, A2 and B1 scenarios [
downscaled climate scenarios were not available at the start of the project. Parameter-specific
monthly climate change anomalies for the study site were derived from the respective grid cell
of the regional climate model COSMO-CLM [
]. The A1B, A2, and B1 scenarios were based
on the global circulation model ECHAM5 and the A1B scenario was also available from the
HadCM3 model [
] (Table 2).
We calculated total deposition of N and S for each site as the sum of throughfall and canopy
exchange. The method is based on a canopy exchange model according to [
] with sodium as
the tracer ion. In order to get long-term deposition of N and S we scaled large-scale modelled
data to the measurements. We used reconstructed deposition from 1880 to 2000 [
deposition values for 2005, 2010, 2020 and 2030 from the latest EMEP model version [
using the current legislation (CLE) scenario with revised Gothenburg Protocol emissions and
the technically maximum feasible emission reduction scenario (MFR). A third deposition
scenario was kept constant after 2010 (B10) (Table 2).
5 / 16
Soil data were taken from soil inventories in 2005 (IM_AT01, IM_AT02) or 2008 (all other
sites). Mineral soils were sampled at depths of 0±5, 5±10, 10±20, 20-(35)40, and (35)40-(50)80
cm from a number (n 3) of soil pits per plot. The organic layer was sampled and analysed
separately. Specimen preparation and analysis followed the methods described in the ICP Forests
monitoring program (http://icp-forests.net/page/icp-forests-manual). The calculation of input
parameters for VSD+ followed the recommendations given in [
] and are described in more
detail in S2 Table (input files in S1 File). Soil solution element concentrations were available
for the 6 evaluation sites and were aggregated to mean annual values for the years 1998 to
Uptake of C, N and base cations as well as above- and belowground litterfall was calculated
using GrowUp, version 1.3.2 [
]. Since both climate change and nitrogen deposition are
supposed to change tree growth and thereby the uptake of N by trees and input via litterfall,
a scaling of these input variables was done comparable to [
]. N and base cation uptake as
well as C and N in litterfall were scaled according to a reference situation, i.e. mean values
between 1970 and 1990 where the core of the forest yield tables [
] was obtained (see details
in S2 File).
For 6 sites detailed data on forest floor vegetation was available for several years between
1990 and 2010 [
]. Forest vegetation at the study sites was recorded on permanent plots using
sub-sampling units which were selected randomly or were distributed in a regular grid.
Permanent plot size was 0.5 ha and sub-sampling unit size was 4 m2. For the ICP Integrated
Monitoring plots IM_AT01 and IM_AT02 8 subsamples per plot were recorded, for ICP Forests sites
10 subsamples were available.
Model calibration and evaluation
The soil model VSD+ was calibrated using measured C and N pools, base saturation and soil
solution chemistry (pH for all sites; NO3- and SO42- for 6 sites) using Markov Chain Monte
Carlo method [
]. Linear regression, mean error and RMSE between measured and modelled
values were calculated for soil C pools, C:N ratios, and soil solution pH, SO42- and NO3-. We
calculated observed HSI by using the frequency of characteristic species in vegetation records
of the sites. Additionally, we compared results from PROPS with the HSI calculated with the
BERN model. BERN differs from PROPS as to the underlying releveÂ data and regarding the
statistics used in deriving niche functions [
]. Linear regressions between the two models
In total, three N and S deposition scenarios (B10, CLE, MFR) and 5 climate scenarios
(including baseline) were modelled for 18 sites resulting in 270 datasets.
First, we applied a two-way ANOVA with F-test statistics to test whether an overall trend
exists over time in temperature (T), precipitation (P), N and S deposition, soil C:N ratio, soil
solution pH and HSI and whether this trend is different between the climate scenarios and the
deposition scenarios. To account for repeated measures over time, the scenarios were nested
into four time slices (2010: average of 2005±15; 2030: 2025±35; 2050: 2045±55; 2100: 2090±
2100) and interactions between climate and deposition scenarios were taken into account.
Second, we used the same ANOVA design for each site. The ANOVA coefficients were used to
determine effect sizes of the climate and deposition scenarios on soil C:N ratio, soil solution
pH and HSI. Differences in effect sizes by 2100 between EU protection habitat types were
tested using one-way ANOVA and Tukey's HSD multi-comparison test. All statistical analyses
were carried out with the package R, version 3.2.3 [
6 / 16
Fig 2. Modelled versus observed Habitat Suitability Index (HSI). Comparison of the modelled Habitat Suitability Index (mean and standard
deviation) using PROPS with A) observed data and B) with modelled HSI from BERN (using the same soil and climate data as PROPS). Six sites
with vegetation records between 1996 and 2007 were used. The 1:1 line is dashed, the regression line is solid.
Model calibration and evaluation
Using linear regression, R2 between modelled and measured HSI was 0.36 but not significant
(p = 0.201) (Fig 2). Mean error was 0.05 and RMSE was 0.12. HSI values as derived from
VSD+ and PROPS were comparable with those derived from VSD+ and BERN (R2 = 0.70,
p = 0.037) but they differed in magnitude, with PROPS giving lower values (mean error was
0.22 and RMSE was 0.26) (Fig 2).
Modelled and measured forest growth achieved an R2 of 0.63 (p<0.001), mean error was
527 kg m-2 and RMSE was 5519. VSD+ could be calibrated to the observed C pool and
C:Nratios resulting into R2 of modelled versus measured values of 0.86 and 0.99, respectively
(p-values < 0.001). The mean error for the C pool was 1.05 kg m-2, RMSE was 2.19. The mean
error for the C:N ratio was 0.13, RMSE was 0.68. As to soil solution composition, only pH
values were significantly related with observations (R2 = 0.57, p<0.001; mean error = 0.41,
RMSE = 1.32). SO42- and NO3- had mean errors of 0.15 and 0.05 eq m-3 and RMSE of 0.21 and
0.07 (S3 File).
Future trends in climate and deposition
Future T and P changed significantly over time in the applied climate scenarios (S4 File).
Whereas T increased by 2.7ÊC on average between 2010 and 2100, P increased slightly towards
2050 and decreased by 76 mm in the year 2100 in comparison to 2010 (Table 2). Only T trends
were significantly different between the climate scenarios. The ECHAM5-A1B, A2 and
HADCM3-A1B scenario showed the highest T increases (3.4±4ÊC by 2100 on average), whereas the
ECHAM5-B1 scenario showed the lowest increase (2.1ÊC). Only in ECHAM5-A2 P decreased
continuously. ECHAM5-A1B and HADCM3-A1B were relatively similar regarding their T
trajectories (the latter showed a faster increase) but differed as to P. Neither increase in T until
7 / 16
2100 nor change in P were significantly different between habitat types (ANOVA p = 0.979 and
In both, the current legislation scenario (CLE) and the maximum feasible reduction
scenario (MFR), N and S deposition decrease significantly over time and significantly differed
from each other (S4 File). Averaged over all sites, N deposition decreased by 2 kg N.ha-1.yr-1
(CLE) and 4 kg N.ha-1.yr-1 (MFR) compared to the deposition in 2010 (Table 2). Differences
in the decrease in N deposition until 2100 between habitat types due to regional differences in
deposition trajectories were marginally significant (ANOVA p = 0.099) due to habitat type
9410 which experienced stronger decrease than 9130 (Tukey's HSD p = 0.053).
Effects of climate, N and S deposition
The average increase of the soil C:N ratio units until 2100 over all scenarios was 3.4 (p<0.001)
(S4 File). When compared to the baseline climate, climate scenarios had an increasing effect
on soil C:N ratios in 82% of the sites and years (2030, 2050 and 2100). HADCM3-A1B had the
strongest effect on C:N ratios (median: 0.72; 90th percentile: 4.29). Climate effects on C:N
ratios by the year 2100 were different between habitat types (ANOVA p<0.001, Fig 3). B10
and CLE deposition scenarios, when compared to the lowest emission scenario (MFR), had
decreasing effects on C:N ratios in 29% of all scenario runs and increasing effects in 69% (S3
Fig 3. Effects of climate change and N deposition scenarios. Effects of climate change (A-C) and N and S
deposition (D-F) on soil C:N ratio, soil pH, and the Habitat Suitability Index (mean and standard deviation of
effects in the year 2100 as derived from 5 different climate change scenarios). Means with different letters are
significant different (Tukey's HSD p < 0.05). Effects are given in the form of ANOVA coefficients describing the
difference between the mean values of all baseline climate model runs and the respective climate change
scenario by the year 2100 and the difference between the mean values of all MFR deposition scenarios and
the respective CLE and B10 deposition scenario by the year 2100 at each site. Positive coefficients represent
increasing, negative coefficients decreasing effects. Note that MFR scenarios have the lowest N deposition.
9110: Luzulo-Fagetum beech forests, 9150: Medio-European limestone beech forests of the
CephalantheroFagion, 9130: Asperulo-Fagetum beech forests, 91G0: Pannonic woods with Quercus petraea and Carpinus
betulus, 9410: Acidophilous Picea forests of the montane to alpine levels (Vaccinio-Piceetea).
8 / 16
Table). The effects of N deposition were an order of magnitude lower than the climate effects
(B10: median = 0.03 and 90th percentile = 0.15). Deposition effects on C:N ratios by the year
2100 were mostly negative and differed between habitat types (ANOVA p<0.001, Fig 3).
The average soil solution pH over all scenarios increased slightly between 2010 and 2100,
but not significantly (p = 0.209) (S4 File). When compared to the baseline climate, climate
scenarios had an increasing effect on soil solution pH in 50% of the sites and years (2030, 2050
and 2100) whereas 22% remained unaffected and in 28% we found a decreasing effect (S3
Table). Climate effects on soil pH by the year 2100 were indifferent between habitat types
(ANOVA p = 0.136, Fig 3). Higher future N deposition had decreasing effects on soil pH in
57% of all model runs and no effect in 42%. Deposition effects on soil pH by the year 2100
were different between habitat types (ANOVA p<0.028, Fig 3). Overall, the effects of different
climate and deposition scenarios on soil solution pH were rather small (10% percentile -0.11±
0.02; 90% percentile: 0±0.18) (S3 Table).
The average HSI over all scenarios decreased significantly from 2010 to 2100 (from 0.2 to
0.14, p<0.001) (S4 File). When compared to the baseline climate, climate scenarios had a
negative effect on HSI in 43% of the sites and years (2030, 2050 and 2100), 21% had a positive effect
and 36% had no effect (S3 Table). Both A1B scenarios had the highest proportion of negative
effects, ECHAM5-B1 the lowest. Positive and negative effects increased over the years so that
only in 11±22% of the climate scenario runs in 2100 no effects were found. With higher
deposition in B10 and CLE deposition scenario as compared to MFR, HSI was negatively affected
in only 15%, positively affected in 45% and 40% of the model runs did not cause an effect. The
effects of different climate and deposition scenarios on HSI were rather small (10% percentile
of all model runs: -0.06 to -0.01; 90% percentile: 0.01±0.04) (S3 Table).
Climate effects on HSI by the year 2100 were indifferent between habitat types (ANOVA
p = 0.255). However, for sites belonging to the EU habitat type 91G0, HSI exhibited the
strongest negative climate effects (Fig 3) resulting from a disproportional number of plant species
with decreasing occurrence probability, most strongly Stellaria holostea L. and Carpinus
betulus (S5 File). Also sites belonging to EU habitat types 9110, 9150, and 9410 HSI experienced
negative climate change effects on average. The plant species experiencing the strongest
negative effects were (in descending order) Linnaea borealis L. and Lonicera caerulea L. (EU habitat
type 9410), Hieracium murorum L. and Luzula luzuloides (Lam.) Dandy & Wilmott. (EU
habitat type 9110), and Mercurialis perennis L. (EU habitat type 9150). In contrast to these
habitats, mean effect on HSI in the EU habitat 9130 was small, and in some cases even positive
(ECHAM5-B2 and HADCM3-A1B) (Fig 3). There, a small number of plant species showed
strongly increasing occurrence probability, particularly in the habitat type 91G0 (Rosa canina
L., Hedera helix L., Quercus petrea). The strongest negative effects were found for Senecio
ovatus (G.Gaertn. et al.) Willd. and Sorbus aucuparia L., the strongest positive effects for Huperzia
selago (L.) Bernh. and Vaccinium myrtillus L. (S5 File).
As mentioned above, HSI showed a positive response to higher N and S deposition in the
B10 and CLE scenario as compared to the MFR scenario. Deposition effects on HSI by the
year 2100 were indifferent between habitat types (ANOVA p = 0.742, Fig 3). The differences in
the occurrence probabilities of plant species between the baseline climate and climate
scenarios was much higher (between -60 and +30%) than the differences between deposition
scenarios. MFR versus B10 scenarios resulted into differences in the range of -4 to +16%. The species
with the strongest increases per habitat type were Hieracium murorum and Vaccinium
myrtillus (EU habitat type 9110), Carpinus betulus and Corylus avellana L. (EU habitat type 91G0),
Milium effusum L. and Senecio ovatus (EU habitat type 9130), Carex sylvatica Huds. and
Athyrium filix-femina (L.) Roth (EU habitat type 9410), and Fagus sylvatica and Mercurialis perennis
(EU habitat type 9150) (S5 File).
9 / 16
Climate change effects on soil chemistry and HSI
According to our climate change scenarios, T increased and P decreased towards 2100 and
these changes did not differ between forest habitat types albeit they differed between scenarios.
When singled out, i.e. calculating differences between the baseline and climate change
scenarios, climate strongly increased the soil C:N ratio. This increase in soil N deficiency was mostly
triggered by N immobilisation in woody biomass as derived from empirical relationships
between climate and forest growth [
]. Interactions of temperature and precipitation were
taken into account insofar as to limit tree growth, particularly in relatively dry areas in eastern
Austria which are likely becoming even more water deficient under expected climate change.
As a consequence, the C:N ratio of oak forest (habitat 91G0) soils in these areas were negatively
affected by climate change.
We expected that these climate changes should directly increase the occurrence probability
of the most thermophilic plant species among the distinctive species of a habitat while the
most cold-tolerant species should decrease. Indeed, the species experiencing the strongest
negative climate effect were cold-tolerant species (e.g. Linnaea borealis, Lonicera caerulea) and
many of those increasing were thermophilic (Rosa canina, Hedera helix). Additionally, and
except for the dry forest habitat 91G0, oligotrophic species should increase while species
preferring sites with higher nutrient availability should decrease. Likewise, oligotrophic species
increased (e.g. Vaccinium myrtillus) but not so in the dry habitat 91G0. There, the
thermophilic species of nutrient rich sites, Stellaria holostea L., increased. This is in accordance with
decreasing soil C:N ratios indicating higher nutrient availability. Species such as Hieracium
murorum and Luzula luzuloides decreased in acidophytic beech dominated forests (habitat
9110). These are typical species of these forests but future N availability may decline to an
extent which lowers their abundance. These changes in species occurrence probabilities
decreased the HSI in all habitats with no significant differences in the magnitude between
them because direct climate effects on the habitat suitability predominated. To our knowledge,
no comparable study exists so far. However, impacts studies have shown the decisive role of
climate for the future of tree species in Austria [
] and neighbouring countries [
N and S deposition effects on soil chemistry and HSI
Contrary to our expectations, the response of the HSI to higher N and S deposition in the B10
and CLE scenario as compared to the MFR scenario was positive. We found a slightly
increasing overall trend in soil solution pH, corroborating observations from the Austrian forest soil
] and other long-term observations across Europe [
29, 30, 65, 66
]. The small
change in soil pH was no surprise since some of the sites are well buffered, S deposition
declined most strongly in the 1980s and is on a rather low level since several years , and N
deposition continued to acidify soils. When singling out deposition effect on soil pH from this
long-term trend, beech dominated forests on acidic soils (9110), were more affected than all
other habitats. These soils particularly acidified during the past period of high S deposition,
and recovery from acidification might be more efficient than in spruce dominated forests. We
note that climate warming can, in addition, increase base cation input to the soil via
accelerated weathering and litter decomposition [
], but significant effects of climate change on
soil pH were only found in some of our study sites. Higher N deposition predominantly led to
increasing soil C:N ratios until 2030 and 2050, but by 2100 most sites experienced negative
effects. The effects were quite small (means < 1) and very likely irrelevant for plant species
occurrence. A decrease in the soil C:N ratio in response to N deposition was found in
10 / 16
numerous studies [
]. However, N deposition may also stimulate plant growth, causing
increases in soil C:N ratios [
]. Most of the study sites were N limited and exposed to
relatively low N deposition, so that the implemented tree growth response to N deposition seems
reasonable. As to plant species responses, we found what we expected: plant species preferring
nutrient-rich sites increased to some extent (e.g. Corylus avellana, Milium effusum, Senecio
ovatus, Carex sylvatica, Athyrium filix-femina, Mercurialis perennis). However, the resulting
increase in the HSI means that additional N improved habitat suitability at these forest sites
which might be due to the fact that N deficiency is still widespread in Austria, because of
historic overuse and acidification during the last part of the 20th century [
the difficulty to explain the increase in HSI, deposition effects on species occurrence
probabilities, and hence the HSI, were much smaller than effects from climate causing negative total
changes in habitat suitability.
Interactions of climate and N deposition
First, we note that the forest sites used for these analyses represent the major forest habitats in
Austria with a wide distribution but do not cover many rarer habitats under conservation
protection. Hence, only one of totally 8 of the EU directive Annex I priority habitats [
] could be
included due to a lack of data. Our results suggests that while climate change will clearly lower
the ªfavourable conservation statusº of these Austrian forest habitats N deposition effects will
be comparably weak. The reasons are twofold. First, N deposition in most of these forests will
not exceed loads at which major changes in the soil chemistry occur, and, secondly, climate
driven increase in N immobilisation in woody biomass will offset soil N enrichment, a result
which is in line with Butler et al. [
]. Our results are also in line with observations showing
that during the last decades European forests have become more nutrient deficient albeit N
deposition was relatively high [
]. They showed that enhanced forest growth due to N
deposition, climate warming and, possibly, CO2 fertilization increased the demand for N and other
nutrients rendering soils more nutrient limited, underlining earlier hypothesis on progressive
N limitation [
]. Indeed, incorporation of nutrient limitation other than N is a clear research
field for the future. Observations are meagre but suggest that in some areas N (and S)
deposition has caused changes in forest floor species composition [
], while soil inventory data
showed increases in pH but, together with needle leaf N concentrations, did not give an
indication of large-scale soil N enrichment .
Measures of habitat suitability, such as the HSI, have the advantage that they are directly
relevant for habitat management and conservation policies [
44, 71, 72
]. According to the
European Union conservation legislation, a protected habitat is considered to have ªfavourable
conservation statusº if ªthe specific structure and functions which are necessary for its
longterm maintenance existº [
]. The HSI is based on predictions from empirical niche functions
of distinctive species in a habitat [
]. While plant species' niche functions have been used
extensively in climate impact assessments and seem to be well defined [
], their usefulness
in studying air pollution effects is just being discussed . Here we applied niche functions
implemented in the newly developed model PROPS, which uses mean annual T, annual total
P, annual N deposition, soil C:N ratio, and the soil solution pH as predictor variables. We were
able to model the soil chemistry in a reasonable way using the dynamic soil model VSD+ [
When comparing the results of PROPS with those from a second plant response model
(BERN), which is based on an independent empirical data set and which is rather different as
to its statistical approach [
], they showed a high correlation. Hence, the niche functions,
11 / 16
which are implemented in the current version of PROPS are, in general, reliable and this is
obvious, because they are statistical representations of observed species occurrence data.
However, on a site scale, and particularly when management comes into play, predictions are
inherently difficult and hard to validate, a fact that has caused considerable discussion in climate
impact studies before [
] as has the applicability of empirical niche functions to perform
future impact assessments [
] Nevertheless, we could show that climate and air pollution
effects on habitat suitability significantly interact yet often in an idiosyncratic way. Owing to
this complexity and the still high uncertainty in present knowledge, adaptation to protect
forest habitat biodiversity, such as defined under the European Habitat Directive, will be
challenging. Dynamic soil-plant models can play an important role as supporting tools to assess
possible future trajectories.
S1 Table. List of all distinctive plant species for the study sites.
S2 Table. Methods used for soil and climate input data for VSD+.
S3 Table. Site- and scenario-specific effects of climate change and N deposition.
S1 File. VSD+ input data.
S2 File. Methods and results for forest growth.
S3 File. Soil model validation.
S4 File. Future changes at the site level.
S5 File. Future changes in the occurrence probability of plant species.
The study was funded by the Austrian Climate Research Program (ACRP5ÐCCN-AdaptÐ
KR12AC5K01308) and by the Austrian Federal Ministry of Agriculture, Forestry. We are
grateful to the data providers of the UNECE-CLRTAP International Cooperative Programmes
ICP Forests and ICP Integrated Monitoring.
Conceptualization: Thomas DirnboÈck, Max Posch, Gert Jan Reinds, Wieger G. W.
Data curation: Thomas DirnboÈck, Ika Djukic, Barbara Kitzler, Johannes Kobler, Max Posch,
Gert Jan Reinds, Angela Schlutow, Franz Starlinger.
Formal analysis: Thomas DirnboÈck.
Funding acquisition: Thomas DirnboÈck.
12 / 16
Investigation: Thomas DirnboÈck. Methodology: Thomas DirnboÈck, Johannes Kobler, Max Posch, Gert Jan Reinds, Wieger G. W. Wamelink.
Project administration: Thomas DirnboÈck.
Software: Janet P. Mol-Dijkstra, Max Posch, Gert Jan Reinds, Angela Schlutow, Wieger G. W.
Writing ± original draft: Thomas DirnboÈck, Wieger G. W. Wamelink.
Writing ± review & editing: Thomas DirnboÈck.
13 / 16
14 / 16
15 / 16
1. Assessment ME . Ecosystems and human well-being: biodiversity synthesis . Washington, DC: World Resources Institute; 2005 .
2. Peñuelas J , Sardans J , Estiarte M , Ogaya R , Carnicer J , Coll M , et al. Evidence of current impact of climate change on life: a walk from genes to the biosphere . Global Change Biology . 2013 ; 19 ( 8 ): 2303 ± 38 . https://doi.org/10.1111/gcb.12143 PMID: 23505157
3. Parmesan C . Ecological and evolutionary responses to recent climate change . Annual Review Ecol Evol Syst . 2006 ; 37 : 637 ± 69 .
4. Bobbink R , Hicks K , Galloway J , Spranger T , Alkemade R , Ashmore M , et al. Global assessment of nitrogen deposition effects on terrestrial plant diversity: a synthesis . Ecological Applications . 2010 Jan; 20 ( 1 ): 30 ± 59 . PMID: 20349829
5. European Commission . Council Directive 92 /43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora ( 1992 ).
6. Posch M , Hettelingh JP , Slootweg J , Reinds GJ . Deriving critical loads based on plant diversity targets . In: Slootweg J , Posch M , Hettelingh JP , Mathijsssen L , editors. Modelling and Mapping the impacts of atmospheric deposition on plant species diversity in Europe . Bilthoven, NL: RIVM; 2014 . p. 41 ± 53 .
7. Bleeker A , Hicks WK , Dentener F , Galloway J , Erisman JW. N deposition as a threat to the World's protected areas under the Convention on Biological Diversity . Environmental Pollution . 2011 ; 159 ( 10 ): 2280 ±8. https://doi.org/10.1016/j.envpol. 2010 . 10 .036 PMID: 21122958
8. Pereira HM , Leadley PW , ProencËa V , Alkemade R , Scharlemann JPW , Fernandez-ManjarreÂs JF , et al. Scenarios for Global Biodiversity in the 21st Century. Science . 2010 ; 330 ( 6010 ): 1496 ± 501 . https://doi. org/10.1126/science.1196624 PMID: 20978282
9. Bellard C , Bertelsmeier C , Leadley P , Thuiller W , Courchamp F. Impacts of climate change on the future of biodiversity . Ecology Letters . 2012 ; 15 ( 4 ): 365 ± 77 . https://doi.org/10.1111/j.1461- 0248 . 2011 . 01736 . x PMID : 22257223
10. Vet R , Artz RS , Carou S , Shaw M , Ro C-U , Aas W , et al. A global assessment of precipitation chemistry and deposition of sulfur, nitrogen, sea salt, base cations, organic acids, acidity and pH, and phosphorus . Atmospheric Environment . 2014 ; 93 :3± 100 .
11. OECD. Climate Change Mitigation: Policies and Progress . Paris: OECD Publishing; 2015 .
12. Maes SL , De Frenne P , Brunet J , de la Peña E , Chabrerie O , Cousins SAO , et al. Effects of enhanced nitrogen inputs and climate warming on a forest understorey plant assessed by transplant experiments along a latitudinal gradient . Plant Ecology . 2014 ; 215 ( 8 ): 899 ± 910 .
13. Engler R , Randin CF , Thuiller W , Dullinger S , Zimmermann NE , ArauÂjo MB , et al. 21st century climate change threatens mountain flora unequally across Europe . Global Change Biology . 2011 ; 17 ( 7 ): 2330 ± 41 .
14. Lenoir J , GeÂgout JC , Marquet PA , de Ruffray P , Brisse H. A significant upward shift in plant species optimum elevation during the 20th century . Science . 2008 ; 320 : 1768 ± 71 . https://doi.org/10.1126/science. 1156831 PMID: 18583610
15. Thuiller W , Lavorel S , ArauÂjo MB , Sykes MT , Prentice IC . Climate change threats to plant diversity in Europe . Proceedings of the National Academy of Sciences . 2005 ; 102 ( 23 ): 8245 ± 50 .
16. Suding KN , Collins SL , Gough L , Clark C , Cleland EE , Gross KL , et al. Functional- and abundancebased mechanisms explain diveristy loss due to N fertilization . Proceedings of the National Academy of Sciences . 2005 ; 102 ( 12 ): 4387 ± 92 .
17. Gilliam FS . Response of the herbaceous layer of forest ecosystems to excess nitrogen deposition . Journal of Ecology . 2006 ; 94 : 1176 ± 91 .
18. Meunier CL , Gundale MJ , SaÂnchez IS , Liess A . Impact of nitrogen deposition on forest and lake food webs in nitrogen-limited environments . Global Change Biology . 2016 ; 22 ( 1 ): 164 ± 79 . https://doi.org/10. 1111/gcb.12967 PMID: 25953197
19. Butterbach-Bahl K , Gundersen P. Nitrogen processes in terrestrial ecosystems . In: Sutton MA , Howard CM , Erisman JW , Billen G , Bleeker A , Grennfelt P , et al., editors. The European Nitrogen Assessment : Cambridge University Press; 2011 . p. 99 ± 124 .
20. Sardans J , Rivas-Ubach A , Peñuelas J. The C : N:P stoichiometry of organisms and ecosystems in a changing world: A review and perspectives . Perspectives in Plant Ecology, Evolution and Systematics . 2012 ; 14 ( 1 ): 33 ± 47 .
21. Cools N , Vesterdal L , De Vos B , Vanguelova E , Hansen K. Tree species is the major factor explaining C:N ratios in European forest soils . Forest Ecology and Management . 2014 ; 311 ( 0 ):3± 16 .
22. Templer PH , Mack MC , Iii FSC , Christenson LM , Compton JE , Crook HD , et al. Sinks for nitrogen inputs in terrestrial ecosystems: a meta-analysis of 15N tracer field studies . Ecology . 2012 ; 93 ( 8 ): 1816 ± 29 . PMID: 22928411
23. Zechmeister-Boltenstern S , Keiblinger KM , Mooshammer M , Peñuelas J , Richter A , Sardans J , et al. The application of ecological stoichiometry to plant±microbial±soil organic matter transformations . Ecological Monographs . 2015 ; 85 ( 2 ): 133 ± 55 .
24. Mulder C , Hettelingh JP , Montanarella L , Pasimeni MR , Posch M , Voigt W , et al. Chemical footprints of anthropogenic nitrogen deposition on recent soil C: N ratios in Europe . Biogeosciences. 2015 ; 12 ( 13 ): 4113 ± 9 .
25. Porter EM , Bowman WD , Clark CM , Compton JE , Pardo LH , Soong JL . Interactive effects of anthropogenic nitrogen enrichment and climate change on terrestrial and aquatic biodiversity . Biogeochemistry . 2013 ; 114 ( 1 ): 93 ± 120 .
26. Butler SM , Melillo JM , Johnson JE , Mohan J , Steudler PA , Lux H , et al. Soil warming alters nitrogen cycling in a New England forest: implications for ecosystem function and structure . Oecologia . 2012 ; 168 ( 3 ): 819 ± 28 . https://doi.org/10.1007/s00442-011 -2133-7 PMID: 21983640
27. McDonnell TC , Belyazid S , Sullivan TJ , Sverdrup H , Bowman WD , Porter EM . Modeled subalpine plant community response to climate change and atmospheric nitrogen deposition in Rocky Mountain National Park , USA. Environmental Pollution. 2014 ; 187 : 55 ± 64 . https://doi.org/10.1016/j.envpol. 2013 . 12 .021 PMID: 24448482
Effect of nitrogen deposition reduction on biodiversity and carbon sequestration . Forest Ecology and Management . 2009 ; 258 ( 8 ): 1774 ± 9 .
29. Akselsson C , Hultberg H , Karlsson PE , Pihl Karlsson G , Hellsten S. Acidification trends in south Swedish forest soils 1986±2008ÐSlow recovery and high sensitivity to sea-salt episodes . Science of The Total Environment . 2013 ; 444 ( 0 ): 271 ± 87 .
30. Cools N , De Vos B. Availability and evaluation of European forest soil monitoring data in the study on the effects of air pollution on forests . iForestÐBiogeosciences and Forestry . 2011 2011- 11 - 03 ; 4 ( 5 ): 205 ± 11 .
31. DirnboÈck T , Grandin U , Bernhardt-RoÈmermann M , Beudert B , Canullo R , Forsius M , et al. Forest floor vegetation response to nitrogen deposition in Europe . Global Change Biology . 2014 ; 20 ( 2 ): 429 ± 40 . https://doi.org/10.1111/gcb.12440 PMID: 24132996
32. Reinecke J , Klemm G , Heinken T. Vegetation change and homogenization of species composition in temperate nutrient deficient Scots pine forests after 45 yr . Journal of Vegetation Science . 2014 ; 25 ( 1 ): 113 ± 21 .
33. Vanhellemont M , Baeten L , Verheyen K. Relating changes in understorey diversity to environmental drivers in an ancient forest in northern Belgium . Plant Ecology and Evolution . 2014 ; 147 ( 1 ): 22 ± 32 .
34. Aherne J , Posch M , Forsius M , Lehtonen A , HaÈrkoÈnen K . Impacts of forest biomass removal on soil nutrient status under climate change: a catchment-based modelling study for Finland . Biogeochemistry . 2011 ; 107 ( 1 ): 471 ± 88 .
35. Gaudio N , Belyazid S , Gendre X , Mansat A , Nicolas M , Rizzetto S , et al. Combined effect of atmospheric nitrogen deposition and climate change on temperate forest soil biogeochemistry: A modeling approach . Ecological Modelling . 2015 ; 306 : 24 ± 34 .
36. De Vries W , Wamelink GWW , Dobben Hv , Kros J , Reinds GJ , Mol-Dijkstra JP , et al. Use of dynamic soil±vegetation models to assess impacts of nitrogen deposition on plant species composition: an overview . Ecological Applications . 2010 ; 20 ( 1 ): 60 ± 79 . PMID: 20349830
37. Belyazid S , Kurz D , Braun S , Sverdrup H , Rihm B , Hettelingh J-P. A dynamic modelling approach for estimating critical loads of nitrogen based on plant community changes under a changing climate . Environmental Pollution . 2011 ; 159 ( 3 ): 789 ± 801 . https://doi.org/10.1016/j.envpol. 2010 . 11 .005 PMID: 21145634
38. Rizzetto S , Belyazid S , GeÂgout J-C , Nicolas M , Alard D , Corcket E , et al. Modelling the impact of climate change and atmospheric N deposition on French forests biodiversity . Environmental Pollution . 2016 ; 213 : 1016 ± 27 . https://doi.org/10.1016/j.envpol. 2015 . 12 .048 PMID: 26809502
39. Schlutow A , DirnboÈck T , Pecka T , Scheuschner T . Use of an empirical model approach for modelling trends of ecological sustainability . In: De Vries W, Hettelingh JP , Posch M , editors. Critical Loads and Dynamic Risk Assessments: Nitrogen, Acidity and Metals in Terrestrial and Aquatic Ecosystems . Dordrecht, NL: Springer; 2015 . p. 381 ± 400 .
40. Bonten LTC , Reinds GJ , Posch M. A model to calculate effects of atmospheric deposition on soil acidification, eutrophication and carbon sequestration . Environmental Modelling & Software . 2016 ; 79 : 75 ± 84 .
41. Reinds GJ , Mol-Dijkstra JP , Bonten L , Wamelink GWW , De Vries W , Posch M. VSD+ PROPS: Recent Developments . In: Slootweg J , Posch M , Hettelingh JP , Mathijssen L , editors. Modelling and Mapping the impacts of atmospheric deposition on plant species diversity in Europe . Bilthoven, NL; 2014 . p. 47 ± 53 .
42. Elith J , Leathwick JR . Species distribution models: ecological explanation and prediction across space and time . Annual Review of Ecology , Evolution, and Systematics . 2009 ; 40 : 677 ± 97 .
43. TuÈxen R. Die heutige potentielle natuÈrliche Vegetation als Gegenstand der Vegetationskartierung . Angewandte Pflanzensoziologie (Stolzenau) . 1956 ; 13 :5± 42 .
44. Rowe EC , Ford AES , Smart SM , Henrys PA , Ashmore MR . Using Qualitative and Quantitative Methods to Choose a Habitat Quality Metric for Air Pollution Policy Evaluation . PLoS ONE . 2016 ; 11 ( 8 ): e0161085. https://doi.org/10.1371/journal.pone. 0161085 PMID: 27557277
45. UNECE . The 1999 Gothenburg Protocol to abate acidification, eutrophication and ground-level ozone . Geneva, Switzerland: UNECE1999 .
46. Neumann M , Schnabel G , GaÈrtner M , Starlinger F , FuÈrst A , Mutsch F , et al. Forest Condition Monitoring in AustriaÐResults of the permanent observation plots (Level II) . Vienna: Austrian Research Centre for Forests 2001 .
47. Mutsch F , Leitgeb E , Hacker R , Amann C , Aust G , Herzberger E , et al. Projekt BioSoil±EuropaÈisches Waldboden-Monitoring ( 2006 /07) Datenband OÈ sterreich. Vienna: Austrian Research Centre for Forests 2013 .
48. Jost G , DirnboÈck T , Grabner M-T , Mirtl M. Nitrogen leaching of two forest ecosystems in a karst watershed . Water Air Soil Pollut . 2011 ; 218 ( 1 ): 633 ± 49 .
49. Coleman K , Jenkins DS . RothC- 26 . 3. A model for the turnover of carbon in soil. Model description and users guide . Harpenden, UK: IACR Rothamsted 2005 .
50. European Commission DG Environment. Interpretation Manual of European Union HabitatsÐEUR28 . Brussles: European Commission DG Environment 2013 .
51. StoÈckle CO , Campbell G , Nelson S. ClimGen Manual . Pullman, WA, USA: Biological Systems Engineering Department, Washington State University; 1999 .
52. IPCC AR4 WG1 . Climate Change 2007 : The Physical Science Basis, Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change . S. S, D. Q , M. M , Z. C, M. M , K.B. A , et al., editors. Cambridge, United Kingdom and New York, NY, USA 2007 .
53. Loibl W , Formayer H , SchoÈner W , Truhetz H , Anders I , Gobiet A , et al. reclip:century 1: Models, Data and GHG-Scenarios, Simulations . Vienna: AIT Austrian Institute of Technology 2011 .
54. Adriaenssens S , Staelens J , Baeten L , Verstraeten A , Boeckx P , Samson R , et al. Influence of canopy budget model approaches on atmospheric deposition estimates to forests . Biogeochemistry . 2013 ; 116 ( 1 ): 215 ± 29 .
55. SchoÈpp W , Posch M , Mylona S , Johansson M. Long-term development of acid deposition (1880±2030) in sensitive freshwater regions in Europe . Hydrol Earth Syst Sci . 2003 ; 7 ( 4 ): 436 ± 46 .
56. Simpson D , Benedictow A , Berge H , BergstroÈm R , Emberson LD , Fagerli H , et al. The EMEP MSC-W chemical transport model and technical description . Atmos Chem Phys . 2012 ; 12 ( 16 ): 7825 ± 65 .
57. De Vries W , Posch M. Modelling the impact of nitrogen deposition, climate change and nutrient limitations on tree carbon sequestration in Europe for the period 1900±2050 . Environmental Pollution . 2011 ; 159 ( 10 ): 2289 ± 99 . https://doi.org/10.1016/j.envpol. 2010 . 11 .023 PMID: 21163561
58. Marschall J . Hilfstafeln fuÈr die Forsteinrichtung . 5 ed. Vienna, Austria: OÈ sterreichischer Agrarverlag; 2011 .
59. Reinds GJ , van Oijen M , Heuvelink GBM , Kros H . Bayesian calibration of the VSD soil acidification model using European forest monitoring data . Geoderma . 2008 ; 146 ( 3 ±4): 475 ± 88 .
60. Team RC . R: A language and environment for statistical computing . Vienna, Austria: R Foundation for Statistical Computing2015.
61. Lexer MJ , HoÈnninger K , Scheifinger H , Matulla C , Groll N , Kromp-Kolb H , et al. The sensitivity of Austrian forests to scenarios of climatic change: a large-scale risk assessment based on a modified gap model and forest inventory data . Forest Ecology and Management . 2002 ; 162 ( 1 ): 53 ± 72 .
62. CH2014 -Impacts. Toward quantitative scenarios of climate change impacts in Switzerland Bern , Switzerland 2014 .
Wohlgemuth T. Climate change and tree responses in Central European forests . Annals of Forest Science . 2015 ; 72 ( 3 ): 285 ± 7 .
64. Jandl R , Smidt S , Mutsch F , FuÈrst A , Zechmeister H , Bauer H , et al. Acidification and Nitrogen eutrophication of Austrian forest soils . Applied and Environmental Soil Science . 2012 ; 2012 : 9 .
65. Matzner E , Zuber T , Lischeid G . Response of soil solution chemistry and solute fluxes to changing deposition rates. Ecological Studies: Biogeochemistry of forested catchments in a changing environment . 2004 ; 172 : 339 ± 60 .
66. Kirk GJD , Bellamy PH , Lark RM . Changes in soil pH across England and Wales in response to decreased acid deposition . Global Change Biology . [https://doi.org/10.1111/j.1365- 2486 . 2009 . 02135 . x]. 2010 ; 16 ( 11 ): 3111 ± 9 .
67. Rowe EC , Jones L , Dise NB , Evans CD , Mills G , Hall J , et al. Metrics for evaluating the ecological benefits of decreased nitrogen deposition . Biological Conservation . 2016 .
68. Jonard M , FuÈrst A , Verstraeten A , Thimonier A , Timmermann V , Potočić N , et al. Tree mineral nutrition is deteriorating in Europe . Global Change Biology . 2015 ; 21 ( 1 ): 418 ± 30 . https://doi.org/10.1111/gcb. 12657 PMID: 24920268
69. Zechmeister HG , DirnboÈck T , HuÈlber K , Mirtl M . Assessing airborne pollution effects on bryophytesÐ Lessons learned through long-term integrated monitoring in Austria . Environmental Pollution . 2007 ; 147 : 696 ± 705 . https://doi.org/10.1016/j.envpol. 2006 . 09 .008 PMID: 17084007
70. HuÈlber K , DirnboÈck T , Kleinbauer I , Willner W , Dullinger S , Karrer G , et al. Long-term impacts of nitrogen and sulphur deposition on forest floor vegetation in the Northern limestone Alps, Austria . Applied Vegetation Science. 2008 ; 11 : 395 ± 404 .
71. Guisan A , Rahbek C. SESAM±a new framework integrating macroecological and species distribution models for predicting spatio-temporal patterns of species assemblages . Journal of Biogeography . 2011 ; 38 ( 8 ): 1433 ± 44 .
72. Delbaere B , Nieto Serradilla A , Snethlage M. BioScore: A tool to assess the impacts of European Community policies on Europe's biodiversity . Tilburg, the Netherlands: ECNC; 2009 .
73. Thuiller W , Lafourcade B , Engler R , ArauÂjo MB . BIOMOD±a platform for ensemble forecasting of species distributions . Ecography . 2009 ; 32 ( 3 ): 369 ± 73 .
74. Rowe EC , Wamelink GWW , Smart SM , Butler A , Henrys PA , van Dobben HF , et al. Field survey based models for exploring nitrogen and acidity effects on plant species diversity and assessing long-term critical loads . In: De Vries W, Hettelingh JP , Posch M , editors. Critical Loads and Dynamic Risk Assessments: Nitrogen, Acidity and Metals in Terrestrial and Aquatic Ecosystems . Dordrecht, NL: Springer; 2015 . p. 297 ± 326 .
75. Randin CF , DirnboÈck T , Dullinger S , Zimmermann NE , Zappa M , Guisan A . Are niche-based species distribution models transferable in space? Journal of Biogeography . 2006 ; 33 ( 10 ): 1689 ± 703 .