Next Article in Journal
Ecosystem Functioning of the Loess Plateau in China from Vegetation Restoration Relied Largely on Climate
Previous Article in Journal
Managing Cone Formation and Leader Growth in Fraser Fir Christmas Tree Plantations with Plant Growth Regulators
Previous Article in Special Issue
The Climate-Growth Relationship between Picea smithiana (Wall.) Boiss. and Abies pindrow (Royle ex D.Don) Royle along the Latitudinal Gradient in Northern Pakistan
 
 
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

The Role of Provenance for the Projected Growth of Juvenile European Beech under Climate Change

1
Karlsruhe Institute of Technology (KIT), Institute of Meteorology and Climate Research—Atmospheric Environmental Research (IMK-IFU), 82467 Garmisch-Partenkirchen, Germany
2
Global Change Research Institute, Czech Academy of Sciences, 616 62 Brno, Czech Republic
3
Faculty of Forestry, Technical University in Zvolen, 960 01 Zvolen, Slovakia
4
Department of Vegetation Ecology, Institute of Botany CAS, 602 00 Brno, Czech Republic
5
Département des Sciences de l’environnement, Université du Québec à Trois-Rivières, Trois-Rivières, QC G8Z 4M3, Canada
6
Institute of Forest Ecology, Slovak Academy of Sciences, 960 01 Zvolen, Slovakia
7
Department of Silviculture and Agroforestry, College of Forestry, KAU, Thrissur 680654, India
8
Administration of Tatra National Park, Tatranska Lomnica, 059 60 Vysoke Tatry, Slovakia
*
Author to whom correspondence should be addressed.
Forests 2023, 14(1), 26; https://doi.org/10.3390/f14010026
Received: 22 November 2022 / Revised: 19 December 2022 / Accepted: 20 December 2022 / Published: 23 December 2022
(This article belongs to the Special Issue Impact of Climate Change on Tree Growth)

Abstract

:
European beech is one of the most common tree species in Europe and is generally suggested to play even more of a prominent role in forestry in the future. It seems to have the potential to partially replace Norway spruce, as it is less sensitive to expected warmer and drier conditions. It is, however, not well known in which regions these new plantings would be particularly favourable and if specific provenances may be better adapted to the new conditions than others. Therefore, we estimated the potential early height growth under climate conditions in 2040–2060 for 20 beech provenances across a region covering the Czech Republic and Slovakia. This Central European region is expected to experience considerably drier and warmer conditions in the future. For this exercise, we implemented a new neural network model developed from height growth information obtained from the open-access BeechCOSTe52 database. The simulations are driven by past and future climate data obtained from the WorldClim database of historical climate data and future climate projections. Simulations revealed that provenances originating from drier regions performed on average significantly better than those from regions with good water supply. Moreover, provenances originating from drier regions had a particularly large advantage in the relatively arid regions of Central Czechia and Southern Slovakia. We can also confirm that all provenances showed a high phenotypic plasticity of height growth across the whole investigated region.

1. Introduction

The global climate change (GCC) affects productivity and resilience of many forest ecosystems across Europe [1,2,3,4]. European beech (Fagus sylvatica L.) is the most abundant tree species in Central Europe and, therefore, an essential target for investigations that elucidate its growth responses under future climate conditions [5,6,7,8]. Generally, adverse effects of decreased precipitation, extreme summer temperature, and an increased frequency of late frosts in spring has been demonstrated [9,10,11]. Drought impacts can be particularly expressed since a large leaf area of European beech leads to relatively high evaporative demands, and the fine root system is expensive to maintain [12,13]. In addition, extreme heat in spring and summer has been shown to decrease assimilation due to depigmentation of beech leaves [14,15,16]. In addition, late frost events as well as drought stress can lead to premature defoliation in European beech, which negatively affects growth in the current and the following vegetation season [17,18]. Thus, large uncertainties exist about the future development of beeches in Central Europe, which may be addressed with scenario simulations that consider future environmental conditions [19]. A simulation study indicated a growth decline between 10–16% in Denmark [20]. Moreover, recent scenario analysis by Martinez del Castillo et al. [21] showed that a reduction of biomass production of European beech forests throughout the whole of Europe can be expected, which would also negatively affect carbon sequestration [22,23]. In the Czech Republic and Slovakia, growth of European beech has already declined during the past 20 years, which was attributed to reduced water availability [24,25,26,27].
In order to avoid or minimise growth reductions, adaptation strategies are needed to increase the resilience of beech under a warmer and drier climate. This includes the application of ‘assisted migration’, which means an intentional translocation of individuals within the natural range of a species [28,29]. Assisted migration assumes that a specific population adapts with time to their local site conditions and will express a specific phenotype via natural selection [30]. Indeed, it has been demonstrated that local adaptations show improved performance for the specific environmental limitations under which they have been developed [31,32]. For practical purposes, common gardens, where multiple provenances are grown together at one site, allow to differentiate between the effects of genotype and environment and enable to select specific populations according to their trait performance [33]. The prevailing hypothesis in provenance research is that southern populations from more xeric and warmer environments are better adapted to warmer and drier conditions in the future [34,35]. In other words, we assume that the hot and dry sites today reflect the conditions of mesic sites in the future. The local adaptation of southern provenances to drought can be then reflected in various morpho-physiological and biochemical traits [36]. A better performance of southern populations under marginal environmental conditions has indeed been observed in many studies [32,37,38,39,40,41], while no particular trade-off between drought resistance and growth has been found [42].
Another important component of population resilience is phenotypic plasticity, which is the range of phenotypes expressed under specific environmental conditions [43]. This can be evaluated by investigations at multiple common garden sites using the ‘ecodistance’ metric introduced by Mátyás [33]. Ecodistance is the difference of an environmental variable between the site of origin and the new location where a provenance has been transferred to. It can be defined with any kind of site characteristic but is typically linked to a climatic variable (e.g., average temperature of the growing season). It has already been successfully used to explain climate-dependent intra-specific growth variability of European beech across Europe [44,45]. Here, we apply several ecodistance-metrics together and link them to the height growth development of provenances using a neural network (NN) approach. The resulting height growth model is then able to project the performance of a specific provenance at any site within the parameterisation space (central Europe here), provided the necessary site information is available.
We hypothesise that beech provenances of southern origin grow better in regions that are expected to experience warmer and drier conditions. In contrast, we expect that provenances from sites that are characterised by a relatively good water supply will suffer most under expected climate change. To evaluate these hypotheses, we use the open-access BeechCOSTe52 database [46], providing information from 39 trial sites and 217 provenances covering the whole distribution of beech in Central Europe to develop the NN-based growth model. The model is then run with a climate scenario running up to 2060, to compare future plasticity and growth of selected provenances, specifically for the regions of the Czech Republic and Slovakia.

2. Materials and Methods

2.1. Phenotypic Data

The data for feeding the NN approach are derived from the open-access BeechCOSTe52 database (version 3.0), which originated from the COST Action E52 [46] available at https://zenodo.org/record/1040664 (accessed on 19 December 2022). The dataset contains phenotypic information of two sets of provenances transferred to common garden sites across Europe. The database includes the longitude and latitude of each provenance’s origin and those of the common gardens they were planted into (Figure 1). The first set of provenances was planted in 1995, and the second set in 1998. In both cases, the height of the trees was measured seven years after the translocation to the new environment. For this analysis, we merged the two provenance sets (1995–2002 and 1998–2005) so that in total, information of 46289 individuals from 227 provenances planted at 13 common gardens sites was derived. IDs of provenances and common gardens are kept identical as given in the BeechCOSTe52 database, available at https://zenodo.org/record/1040664 (accessed on 19 December 2022). The specific height development, seven years after translocation, is presented for all provenances in the Supplementary Data file and visualised in Supplementary Figure S1.

2.2. Climatic Data

The climate information at the specific locations of provenance origin, as well as for the common garden sites were downloaded from the WorldClim v2.1 database [47], available at https://www.worldclim.org/data/worldclim21.html (accessed on 19 December 2022). Since the earliest historical weather records in this database are from 1960, we derived the climate for the sites of provenance origin from the period 1961–1990. Use the climate period immediately before seed collection is assumed to deliver the best available estimates for characterising the conditions under which the genotypes developed before being transferred to common garden sites [45]. The environmental conditions that characterise the test sites at which the provenances were growing after the transfer were calculated as mean values for the period from planting (1995/1998) to measurement (2002/2005). The temperature and precipitation distribution at the sites of provenance origin, as well as at the common garden sites, is presented in Figure 2. The ecodistances for all climatic parameters were then calculated as the difference between the respective metric at the provenance origin and that at the common garden site [33]. Historical climate data and ecodistances for all sites and provenances are given in the in Supplementary Materials.
For the scenario calculations (2041–2060), we used the results from the ACCESS-ESM1–5 CMIP6 [48] ensemble simulation (mild Shared Socio-economic Pathways, SSP245, RCP4.5) also available from the WorldClim v2.1 database [47]. The data were then filtered at 15 min spatial resolution for an area within the Czech Republic and Slovakia. For the scenario simulation, the same set of ecodistances as before was calculated as difference between the respective climate metric at the site of provenance origin (1961–1990) and the climate at each grid point within the target region (2040–2060).

2.3. Model Development

To integrate various ecodistance values into a generally applicable site-dependent growth model, we used an artificial neural network (feed-forward back propagation, FFBPNN) approach. Neural networks are extremely useful for solving problems involving multiple influences and complex interactions, such as those given by the interplay of various climate conditions and provenances [49,50]. NNs are based on a general approximation theorem [51] and consist of so-called layers, which are collections of processing nodes (neurons) fully connected with the nodes in other layers but not within their own layer. These layers can be differentiated into input, output, and one or more hidden layers in between. It has been shown that systems with only one hidden layer can already approximate any continuous function, but usually, more than one is used. The reason is that with any additional interaction, a single hidden layer increases exponentially and thus becomes very large when capturing complex relationships. If datasets are small, simple networks with only a few layers and neurons often perform best [52].
Therefore, we used a NN model with three hidden layers consisting of 36, 18, and 9 neurons in each layer (Figure S2). The input layer initially considers 47 variables (climatic characteristics, ecodistance, longitude, latitude, altitude) described in Supplementary Table S1. All climatic influences were centred and normalised to ensure that every influence was equally weighted. In order to reduce the number of variables used in the base line model, we defined so-called hyper parameters that needed to be determined before the training of the final neural network. These hyper parameters determinate the overall structure of the neural network. The selection of hyper parameters values directly affects the performance and generalisation capability of a network. For optimum hyper parameter selection, grid and manual search are the most commonly applied strategies, which both were used here. First, the NN was run with the following setting: activation function was set to rectified linear unit (RELU) and Kernel initialiser was set to normal, learning rate was set to 0.01, number of epochs was set to 500, and batch size was set to 200. Then, the package Dalex, freely available in Phyton libraries, was used to assess the variable importance of investigated variables as predictors. Therefore, a script in Python was developed, combining a 10-fold cross-validation with a hyper parameter search. The best performance was achieved when seven predictors were used according to coefficient of determination (R2), mean square error (MSE), root mean square error (RMSE), and mean absolute deviation (MAD). After identifying and optimising the number of predictors fed into the input layer, the selected hyper parameters were tuned (grid searched). Because the grid search was computationally demanding, an optimisation was carried out per part by considering a combination of grids of two, and three hyper parameters. The batch size was set to 10, 20, 30, 60, 120, and 200, with the number of epochs: 50, 100, 200, 300, 500; drop-out rate was tested with the values: 1, 0.9, 0.8, 0.7, 0.6; learning rate with 0.001, 0.01, 0.05, 0.1; number of hidden layers: 1, 2, 3, 4; number of neurons in a hidden layer: 4, 8, 16, 32, 6. The applied activation functions were: RELU, exponential linear unit (ELU), softmax, sigmoid, linear and tanh; and the kernel initialisation: uniform, normal, and zero. Finally, the RELU function showed the best performance, according to mean absolute error, which was determined by stochastic gradient descent.
For calibration and evaluation, all data were divided into two subsets. The model was trained with 66% of the provenance–ecodistance relationships, while the performance was tested with the remaining subset of data (34% of the total). The analysis was run in Python (3.10, Python Software Foundation, Wilmington, Delaware, USA), using the libraries Pandas 1.51, NumPy 1.23.0 scikit-learn 1.1.3, Keras 2.11.0 and TensorFlow 2.11.0. The final model was run in R (Version 4.2.1, Vienna, Austria) and passed to Dalex 2.4.2 package to assess the ultimate variable importance of the seven predictors and partial dependence plots.

2.4. Statistical Analysis

All statistical analyses were conducted in R 4.2.1 software (R Core Team, Vienna, Austria). ANOVA assumptions of normal distribution of height growth were tested by the Shapiro–Wilk test, and homoscedasticity between months was tested by the Bartlett’s test. Afterwards, the statistical differences in historical seedlings’ height between provenances and common garden sites were tested by two-way ANOVA (initial dataset). The comparison of provenance height simulation values of specific climatic groups was analysed by Tukey’s HSD post-hoc test (future growth data).

3. Results

3.1. Climatic Variable Importance in Model

The European beech provenances showed high intraspecific variability of height growth among the common garden sites (Figure S1). Height differences were significant between the provenances within a common garden site, as well as for the same provenance at the different sites (Table 1). The final NN model showed the R2 of 77% and mean absolute error of 71.25 cm for tree height prediction among the provenances and sites (Figure 3). The sensitivity analysis revealed that out of all the best explanatory values (Figure 4), three were climatic variables from the sites of growth: mean temperature of wettest quarter of the year (Bio8), precipitation of driest month (Bio14), and precipitation of warmest quarter of the year (Bio18). The other four variables describe a particular ecodistance between the site of origin and the site of growth: mean temperature of wettest quarter of the year (Bio8 ED), annual precipitation sum (Bio12ED), precipitation of the wettest month (Bio13ED), and precipitation of warmest quarter of the year (Bio18ED).
The height growth responses in relation to these seven variables, which are used as independent determinants, are shown in Figure 4. They suggest that provenances performed better at warmer environments (Bio8) and sites with higher precipitation during the driest month (Bio14). In addition, provenances showed better height growth at sites warmer than the locations of origin (Bio8ED), or had higher annual precipitation (Bio12ED). Interestingly, translocation to sites with more unevenly distributed rainfall, as expressed by the precipitation during the wettest month of the year (Bio13ED), as well as during the warmest quarter of the year (Bio18ED), had a negative impact. The ecodistance parameters had generally greater importance in the model than the climatic parameters of the original provenance sites.

3.2. Model Climatic Limits and Provenance Grouping

The potential responses of the provenances to future climate conditions were analysed based on the SSP245/RCP4.5 scenario, simulating the height growth for 2041–2060 throughout the Czech Republic and Slovakia at 15 min spatial distribution. Other SSP (higher temperature) scenarios and later time periods exceeded the climatic boundary conditions in which the model has been developed (see Table 2) and were thus not used for the analysis. For simplicity reasons, we grouped all provenances according to their climate of origin into hot and wet, cold and wet, hot and dry, and cold and dry. The threshold for dry vs. wet characterisation was annual precipitation of 800 mm, and the threshold for hot vs. cold characterisation was annual mean temperature of 7 °C. Then we randomly selected five provenances out of each group and simulated their height response for each of the grids within the geographical constraints of the Czecho–Slovak region (Figure 5). The climatic characteristics of the 20 chosen provenances are presented in Table 3.

3.3. Spatial Height Growth Predictions

According to the model, all groups of provenances show a high phenotypic plasticity within the analysed region. The best growth is predicted to occur in the mountainous regions of Central and North Slovakia and the Jeseníky Mountains of Czech Republic. The least height growth has been simulated for southern Slovakia, in the area of the Hungarian border and Central Czechia (Figure 5). The average height within each group of origin (given with standard errors) was 421 ± 2.1 cm for the hot–dry group, 368 ± 4.82 cm for the hot–wet group, 384 ± 1.74 cm for the cold–dry group, and 379 ± 2.92 cm for the cold–wet group (Figure 6). Overall, the provenances from the hot–dry environment show the highest height growth across the whole Czecho–Slovak region. The Austrian provenance AU35 from a high altitude, cold and wet location showed worst height growth performance across provenances overall. Moreover, provenances from the cold–dry environment showed the smallest variability and the ones from the hot–wet environment showed the greatest variability (Figure 6).

4. Discussion

The scenario simulations showed that the mild SSP245/RCP4.5 climate change scenario particularly favours the growth of beech provenances from hot–dry and cold–dry environments. Provenances from more humid environments performed significantly worse than provenances originating from drier regions. The growth for all groups seems to be retarded around southern Slovakia, likely due to the relatively low rainfall and, thus, greater drought risk in this region. Best height growth was predicted for high altitude locations for all provenances. The historical increase in European beech growth at higher altitudes can be attributed to the extended growing season [53]. The spatial distribution of height growth from our simulation generally corresponds to the BAI estimates for beech under a mild scenario for the same period, reported recently by Martinez del Castillo et al. [21]. Moreover, locations with the reduced height growth match with the climate change hot-spots identified by Hlásny et al. [54].

4.1. Ecodistance as a Predictor of Provenance Response

Ecodistance captures the climatic difference between the region of provenance and the new location to which it is transferred. The term was coined by Mátyás [33] and has been widely used in provenance research. Ecodistance variables that indicate the aridity of a site already showed great explanatory power for the interpretation of intra-specific variability of vitality [55], mortality [56], and growth [57] of European beech provenances.
This is in accordance with findings that also investigated beech provenances and demonstrated impacts of both drought and temperature ecodistances on stomatal and leaf morphology [58] and of a drought-related ecodistance on the antioxidant system efficiency [59]. In addition, aridity-related ecodistance also showed great explanatory power regarding the intra-specific variability of root growth dynamics [60]. The results indicate that European beech can acclimate (water uptake and retention) to different climatic conditions as induced by geographic transfer (which can be seen as analogy to climate change). Similar results have also been achieved for other tree species such as oak [61], Norway spruce [62,63] and Scots pine [45].
The application of ecodistance metrics has been used before to explain the intra-specific variability of European beech growth within common garden sites [44,45,64]. However, it is the first time here that it is used for scenario analysis where the ecodistance metrics are calculated from a regionally differentiated future climate. Using a NN model is also a new approach, which elaborates the methodology of Mátyás et al. [55], who conducted mono-dimensional extrapolations of beech height growth based on the ecodistance of Forest Aridity. The relatively weak relationship they found inspired the development of a more complex multi-factorial neural network. The resulting model based on four normalised ecodistance indices and additional three site climate variables is now able to explain 77% of total height variability across provenances and common garden sites. The combination of climate-related ecodistance values can thus certainly be viewed as a powerful tool for the explanation of provenance phenotypic transfer responses.

4.2. Limitations of the Study

It should be noted that the model does not account for soil conditions, which are difficult to define for a region (of origin) and often not available for a target site (at least for regionally distributed simulations). Given the dependence of root systems on soil depth and of potential water uptake from soil texture, including such factors will likely have an additional explanatory impact on provenance development [65]. Additionally, growth limitation due to nutrition constraints can have a significant impact on the accuracy of the presented results [66], although nutrition and water availability are also related to a certain degree [67]. Another influencing factor that we miss is the difference in past and future atmospheric CO2 concentration, which may decrease the impact of aridity and thus mitigate future impacts of drought and high temperatures [68,69]. This is an effect difficult to include in models that are based on statistical performance of the past because future CO2 concentrations have never been experienced before. As another influencing factor, the topography of the locations was not accounted for, since slope and elevation can influence water and nutrient availability beyond what can be estimated from soil texture and depth. However, topography is generally closely related to site climate conditions. It is also not possible to consider the impacts of topography if averaged over a 27.75 km grid which was the spatial resolution in our simulations.
We admit that even though various aspects of tree growth were included, a NN model does not account for any physiological processes or process knowledge. It rather represents a black box that produces output from input based on the statistical relationships of the past. Therefore, it should not be applied outside the range of observations that have been used to calibrate and evaluate the model, which is the reason why we selected a climate scenario that does not reach far into the future, as well as a very specific area, which is close to the region the model has been developed for. Within this domain of climatic constraints, however, we can assume that the model provides reasonable and meaningful results.

4.3. High Phenotypic Plasticity of European Beech

Populations with higher phenotypic plasticity are generally viewed positively as they can adapt to higher environmental variability and thus can minimise the risk of mortality [70]. The distribution of height growth among provenance groups in our study shows great variability (100 to 500 cm growth within a 7-year period), but similar average growth, which is around 388 cm. These results suggest that European beech is a highly plastic species, especially as these groups originated from contrasting environments. Indeed, this is supported by various studies. For example, Müller et al. [71] showed a high height growth plasticity for European beech across common garden sites all over Germany. Similarly, a study concentrating on climate conditions of Western and Central Europe also investigating trees at common gardens confirms this highly plastic growth responses [45], which is furthermore corroborated by various other provenance observations [33,72,73]. A high intra-specific variability of European beech provenances has been observed not only for growth but also for phenology [74,75], stomatal and leaf morphology [57,76,77,78], root architecture [59], xylem embolism resistance [79,80,81], xylem hydraulic conductivity [82], gas-exchange [41], photochemistry [83], δ13C partitioning [84,85], and establishment success [86]. Overall, the broad range of intra-specific variability of European beech across the traits suggests a great adaptation potential under global climate change.

4.4. Drought and Temperature Limited Growth

Despite its high plasticity, the susceptibility of European beech to more extreme heat and drought events or generally drier conditions is highly disputed. A recent meta-analysis by Leuschner et al. [12] concluded that European beech already shows significant growth reductions in large parts of its distribution range. From a regional perspective, it seems that, in particular, the populations at low altitudes are suffering from decreased precipitation, especially during spring [87]. In fact, beech growth in Slovakia has been declining since the year 2000 [26,27], and a similar trend has also been observed in the Czech Republic [24,25]. The dependency of growth on water availability is also supported for close-by regions such as Hungary [44]. The observed growth reductions in European beech have been attributed to recent increases in summer temperatures [88], decreases in summer precipitation [89], or decreases in relative air humidity [90]. In combination, these findings indicate that the reason for growth decline is an increase in overall aridity [91]. However, at sites where aridity is not increasing or water supply is at least still sufficiently available, beech trees are showing increasing growth trends. This is also confirmed for the target region in this study [92,93].
The simulations reflect these observations well. The retarded growth has been simulated particularly in southern Slovakia, closely related to the greater aridity that emerges during the 2040–2060 period from lower precipitation and greater evaporation demand. In contrast, the highest predicted growth occurred in mountainous areas of the Czech Republic and Slovakia, which are expected to be relatively well water supplied in the upcoming years. This increase is likely linked to the consideration of the temperature of the wettest time of the year in model, which is usually the spring period. Growth increases originating from higher spring temperatures are caused by the relation to bud burst, which happens sooner each year, elongating the growing period [94,95]. The spring temperature has been observed as a main explanatory factor for the growth variability of European beech in Central Europe [25]. However, this elongation can also increase the sensitivity to late frosts [17,96], particularly at high elevations. If there are more frequent frosts after bud burst, it can cause retardation of growth since the lost foliage has to be replaced with carbon not available anymore for stem growth [97]. The mixed effects of increasing temperature and aridity can have complex effects on tree growth, as viewed in the sensitivity analysis of our NN model.
Overall, the inclusion of various indices enabled not only the responses at specific sites to be described with higher precision but also coverage of the spatial variability that originates from different dominating influences at different sites. In addition, the regional analysis also enabled to highlight differences in growth responses between provenances, i.e., the better performance of plants originating from drier regions, which are better adapted to the expected future conditions.

5. Conclusions

Our results indicate that growth of young European beech trees will be retarded at drier locations of Southern Slovakia and Central Czechia within 2040–2060 period. Overall, the use of provenances originating from Mediterranean regions, and thus generally better adapted to warmer and drier conditions, will be probably advantageous. The exception is in mountainous areas where drought impacts will not be as pronounced. Our model suggests that precipitation and spring temperatures will be prominent drivers of European beech growth in the future in Central Europe, indicating an increasing significance of drought.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/f14010026/s1, Figure S1: Variability of height among provenance populations at the 13 common garden sites used for NN development; Figure S2: Final architecture of the neural network model. Input layer has seven neurons (climate variables and ecodistances), which are connected to three hidden layers (36, 18, and 9 neurons) and lead to output height. Each line represents a linear or non-linear function between the two connected neurons; Table S1: Initial explanatory variables used for construction of neural network model; Table S2: Full country name of provenance abbreviations used in Table 3.

Author Contributions

P.P. and P.F.J. conceived the research idea and hypotheses. D.G. contributed his data from two Slovak sites to the original BeechCOSTe52 database. D.K. and A.S.K. were in charge of the data acquisition and management. P.P. and P.F.J. conducted the data analysis, simulation and interpreted the results. A.P.-P., M.M., L.J.L., H.D. and P.P. wrote the first draft, which was elaborated by R.G. and D.G., and modified by all co-authors. All authors have read and agreed to the published version of the manuscript.

Funding

This article is based on work from the COST Action CA19128—Pan-European Network for Climate Adaptive Forest Restoration and Reforestation (PEN-CAFoRR), supported by COST (European Cooperation in Science and Technology). The provenance experiment has been established by the European Network for the Evaluation of the Genetic Resources of Beech for Appropriate Use in Sustainable Forestry Management (AIR3-CT94-2091) under the coordination of H.-J. Muhs and G. von Wühlisch. Collection of phenotypic data was accomplished within the COST Action E52: Evaluation of Beech Genetic Resources for Sustainable Forestry. P.P. was supported by the Ministry of Education, Youth and Sports of Czech Republic within the CzeCOS program, grant number LM2018123. D.K., A.S.K. and D.G. were supported by a grant of the Slovak Research and Development Agency no. APVV-21-0270. M.M., A.S.K. and D.K. were supported by the Slovak Grant Agency for Science—grant no. VEGA 1/0535/20.

Data Availability Statement

All data used within this study were open access and are available at respective repositories stated within the article. Climate and ecodistance data used for this study are included in the attached data file in in Supplementary Materials.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Dyderski, M.K.; Paź, S.; Frelich, L.E.; Jagodziński, A.M. How much does climate change threaten European forest tree species distributions? Glob. Chang. Biol. 2018, 24, 1150–1163. [Google Scholar] [CrossRef] [PubMed]
  2. Schuldt, B.; Buras, A.; Arend, M.; Vitasse, Y.; Beierkuhnlein, C.; Damm, A.; Gharun, M.; Grams, T.E.E.; Hauck, M.; Hajek, P.; et al. A first assessment of the impact of the extreme 2018 summer drought on Central European forests. Basic Appl. Ecol. 2020, 45, 86–103. [Google Scholar] [CrossRef]
  3. Bandh, S.A.; Shafi, S.; Peerzada, M.; Rehman, T.; Bashir, S.; Wani, S.A.; Dar, R. Multidimensional analysis of global climate change: A review. Environ. Sci. Pollut. Res. 2021, 28, 24872–24888. [Google Scholar] [CrossRef] [PubMed]
  4. Szatniewska, J.; Zavadilova, I.; Nezval, O.; Krejza, J.; Petrik, P.; Čater, M.; Stojanović, M. Species-specific growth and transpiration response to changing environmental conditions in floodplain forest. For. Ecol. Manag. 2022, 516, 120248. [Google Scholar] [CrossRef]
  5. Gessler, A.; Keitel, C.; Kreuzwieser, J.; Matyssek, R.; Seiler, W.; Rennenberg, H. Potential risks for European beech (Fagus sylvatica L.) in a changing climate. Trees 2006, 21, 1–11. [Google Scholar] [CrossRef]
  6. Etzold, S.; Ziemińska, K.; Rohner, B.; Bottero, A.; Bose, A.K.; Ruehr, N.K.; Zingg, A.; Rigling, A. One Century of Forest Monitoring Data in Switzerland Reveals Species- and Site-Specific Trends of Climate-Induced Tree Mortality. Front. Plant Sci. 2019, 10, 307. [Google Scholar] [CrossRef]
  7. Pretzsch, H.; del Río, M.; Grote, R.; Klemmt, H.-J.; Ordóñez, C.; Oviedo, F.B. Tracing Drought Effects from the Tree to the Stand Growth in Temperate and Mediterranean Forests: Insights and Consequences for Forest Ecology and Management. Eur. J. Forest. Res. 2022, 141, 727–751. [Google Scholar] [CrossRef]
  8. Hlásny, T.; Mátyás, C.; Seidl, R.; Kulla, L.; Merganičová, K.; Trombik, J.; Dobor, L.; Barcza, Z.; Konôpka, B. Climate change increases the drought risk in Central European forests: What are the options for adaptation? For. J. 2014, 60, 5–18. [Google Scholar] [CrossRef]
  9. Rita, A.; Camarero, J.J.; Nolè, A.; Borghetti, M.; Brunetti, M.; Pergola, N.; Serio, C.; Vicente-Serrano, S.M.; Tramutoli, V.; Ripullone, F. The impact of drought spells on forests depends on site conditions: The case of 2017 summer heat wave in southern Europe. Glob. Chang. Biol. 2019, 26, 851–863. [Google Scholar] [CrossRef]
  10. Vitasse, Y.; Bottero, A.; Cailleret, M.; Bigler, C.; Fonti, P.; Gessler, A.; Lévesque, M.; Rohner, B.; Weber, P.; Rigling, A.; et al. Contrasting resistance and resilience to extreme drought and late spring frost in five major European tree species. Glob. Chang. Biol. 2019, 25, 3781–3792. [Google Scholar] [CrossRef]
  11. Leuschner, C. Drought response of European beech (Fagus sylvatica L.)—A review. Perspect. Plant Ecol. Evol. Syst. 2020, 47, 125576. [Google Scholar] [CrossRef]
  12. Petrík, P.; Zavadilová, I.; Šigut, L.; Kowalska, N.; Petek-Petrik, A.; Szatniewska, J.; Jocher, G.; Pavelka, M. Impact of Environmental Conditions and Seasonality on Ecosystem Transpiration and Evapotranspiration Partitioning (T/ET Ratio) of Pure European Beech Forest. Water 2022, 14, 3015. [Google Scholar] [CrossRef]
  13. Bigler, C.; Vitasse, Y. Premature leaf discoloration of European deciduous trees is caused by drought and heat in late spring and cold spells in early fall. Agric. For. Meteorol. 2021, 307, 108492. [Google Scholar] [CrossRef]
  14. Húdoková, H.; Petrik, P.; Petek-Petrik, A.; Konôpková, A.; Leštianska, A.; Střelcová, K.; Kmeť, J.; Kurjak, D. Heat-stress response of photosystem II in five ecologically important tree species of European temperate forests. Biologia 2022, 77, 671–680. [Google Scholar] [CrossRef]
  15. Kunert, N.; Hajek, P. Shade-tolerant temperate broad-leaved trees are more sensitive to thermal stress than light-demanding species during a moderate heatwave. Trees For. People 2022, 9, 100282. [Google Scholar] [CrossRef]
  16. Sangüesa-Barreda, G.; Di Filippo, A.; Piovesan, G.; Rozas, V.; Di Fiore, L.; García-Hidalgo, M.; García-Cervigón, A.I.; Muñoz-Garachana, D.; Baliva, M.; Olano, J.M. Warmer springs have increased the frequency and extension of late-frost defoliations in southern European beech forests. Sci. Total Environ. 2021, 775, 145860. [Google Scholar] [CrossRef]
  17. Arend, M.; Link, R.M.; Zahnd, C.; Hoch, G.; Schuldt, B.; Kahmen, A. Lack of hydraulic recovery as a cause of post-drought foliage reduction and canopy decline in European beech. New Phytol. 2022, 234, 1195–1205. [Google Scholar] [CrossRef]
  18. Sperlich, D.; Nadal-Sala, D.; Gracia, C.; Kreuzwieser, J.; Hanewinkel, M.; Yousefpour, R. Gains or Losses in Forest Productivity under Climate Change? The Uncertainty of CO2 Fertilization and Climate Effects. Climate 2020, 8, 141. [Google Scholar] [CrossRef]
  19. Huang, W.; Fonti, P.; Larsen, J.B.; Ræbild, A.; Callesen, I.; Pedersen, N.B.; Hansen, J.K. Projecting tree-growth responses into future climate: A study case from a Danish-wide common garden. Agric. For. Meteorol. 2017, 247, 240–251. [Google Scholar] [CrossRef]
  20. del Castillo, E.M.; Zang, C.S.; Buras, A.; Hacket-Pain, A.; Esper, J.; Serrano-Notivoli, R.; Hartl, C.; Weigel, R.; Klesse, S.; de Dios, V.R.; et al. Climate-change-driven growth decline of European beech forests. Commun. Biol. 2022, 5, 163. [Google Scholar] [CrossRef]
  21. Arend, M.; Sever, K.; Pflug, E.; Gessler, A.; Schaub, M. Seasonal photosynthetic response of European beech to severe summer drought: Limitation, recovery and post-drought stimulation. Agric. For. Meteorol. 2016, 220, 83–89. [Google Scholar] [CrossRef]
  22. Brunn, M.; Hafner, B.D.; Zwetsloot, M.J.; Weikl, F.; Pritsch, K.; Hikino, K.; Ruehr, N.K.; Sayer, E.J.; Bauerle, T.L. Carbon allocation to root exudates is maintained in mature temperate tree species under drought. New Phytol. 2022, 235, 965–977. [Google Scholar] [CrossRef] [PubMed]
  23. Kolář, T.; Čermák, P.; Trnka, M.; Žid, T.; Rybníček, M. Temporal changes in the climate sensitivity of Norway spruce and European beech along an elevation gradient in Central Europe. Agric. For. Meteorol. 2017, 239, 24–33. [Google Scholar] [CrossRef]
  24. Šimůnek, V.; Vacek, Z.; Vacek, S.; Králíček, I.; Vančura, K. Growth Variability of European Beech (Fagus sylvatica L.) Natural Forests: Dendroclimatic Study from Krkonoše National Park. Cent. Eur. For. J. 2019, 65, 92–102. [Google Scholar] [CrossRef]
  25. Bosela, M.; Štefančík, I.; Petráš, R.; Vacek, S. The effects of climate warming on the growth of European beech forests depend critically on thinning strategy and site productivity. Agric. For. Meteorol. 2016, 222, 21–31. [Google Scholar] [CrossRef]
  26. Bosela, M.; Lukac, M.; Castagneri, D.; Sedmák, R.; Biber, P.; Carrer, M.; Konôpka, B.; Nola, P.; Nagel, T.A.; Popa, I.; et al. Contrasting effects of environmental change on the radial growth of co-occurring beech and fir trees across Europe. Sci. Total Environ. 2017, 615, 1460–1469. [Google Scholar] [CrossRef]
  27. Williams, M.I.; Dumroese, R.K. Preparing for Climate Change: Forestry and Assisted Migration. J. For. 2013, 111, 287–297. [Google Scholar] [CrossRef]
  28. Gömöry, D.; Krajmerová, D.; Hrivnák, M.; Longauer, R. Assisted Migration vs. Close-to-Nature Forestry: What Are the Prospects for Tree Populations under Climate Change? Cent. Eur. For. J. 2020, 66, 63–70. [Google Scholar] [CrossRef]
  29. Lortie, C.J.; Hierro, J.L. A Synthesis of Local Adaptation to Climate through Reciprocal Common Gardens. J. Ecol. 2022, 110, 1015–1021. [Google Scholar] [CrossRef]
  30. Konôpková, A.; Pšidová, E.; Kurjak, D.; Stojnić, S.; Petrík, P.; Fleischer, P., Jr.; Kučerová, J.; Ježík, M.; Petek, A.; Gömöry, D.; et al. Photosynthetic Performance of Silver Fir (Abies alba) of Different Origins under Suboptimal Growing Conditions. Funct. Plant Biol. 2020, 47, 1007. [Google Scholar] [CrossRef]
  31. Ćelepirović, N.; Bogunović, S.; Dounavi, A.; Netzer, F.; Eiblmeier, M.; Dannenmann, M.; Rehschuh, S.; Rennenberg, H.; Ivanković, M. Phosphorus Nutrition and Water Relations of European Beech (Fagus sylvatica L.) Saplings Are Determined by Plant Origin. Forests 2022, 13, 1683. [Google Scholar] [CrossRef]
  32. Mátyás, C. Climatic Adaptation of Trees: Rediscovering Provenance Tests. Euphytica 1996, 92, 45–54. [Google Scholar] [CrossRef]
  33. Guo, X.; Khare, S.; Silvestro, R.; Huang, J.; Sylvain, J.-D.; Delagrange, S.; Rossi, S. Minimum Spring Temperatures at the Provenance Origin Drive Leaf Phenology in Sugar Maple Populations. Tree Physiol. 2020, 40, 1639–1647. [Google Scholar] [CrossRef] [PubMed]
  34. Lu, P.; Parker, W.C.; Colombo, S.J.; Man, R.; Colombo, S. Restructuring tree provenance test data to conform to reciprocal transplant experiments for detecting local adaptation. J. Appl. Ecol. 2016, 53, 1088–1097. [Google Scholar] [CrossRef]
  35. Mukarram, M.; Choudhary, S.; Kurjak, D.; Petek, A.; Khan, M.M.A. Drought: Sensing, Signalling, Effects and Tolerance in Higher Plants. Physiol. Plant. 2021, 172, 1291–1300. [Google Scholar] [CrossRef]
  36. Robson, T.M.; Rasztovits, E.; Aphalo, P.J.; Alia, R.; Aranda, I. Flushing Phenology and Fitness of European Beech (Fagus sylvatica L.) Provenances from a Trial in La Rioja, Spain, Segregate According to Their Climate of Origin. Agric. For. Meteorol. 2013, 180, 76–85. [Google Scholar] [CrossRef]
  37. Gömöry, D.; Ditmarová, Ľ.; Hrivnák, M.; Jamnická, G.; Kmeť, J.; Krajmerová, D.; Kurjak, D. Differentiation in Phenological and Physiological Traits in European Beech (Fagus sylvatica L.). Eur. J. For. Res. 2015, 134, 1075–1085. [Google Scholar] [CrossRef]
  38. Hajek, P.; Kurjak, D.; von Wühlisch, G.; Delzon, S.; Schuldt, B. Intraspecific Variation in Wood Anatomical, Hydraulic, and Foliar Traits in Ten European Beech Provenances Differing in Growth Yield. Front. Plant Sci. 2016, 7, 791. [Google Scholar] [CrossRef]
  39. Knutzen, F.; Meier, I.C.; Leuschner, C. Does Reduced Precipitation Trigger Physiological and Morphological Drought Adaptations in European Beech (Fagus sylvatica L.)? Comparing Provenances across a Precipitation Gradient. Tree Physiol. 2015, 35, 949–963. [Google Scholar] [CrossRef]
  40. Wang, F.; Israel, D.; Ramírez-Valiente, J.-A.; Sánchez-Gómez, D.; Aranda, I.; Aphalo, P.J.; Robson, T.M. Seedlings from Marginal and Core Populations of European Beech (Fagus sylvatica L.) Respond Differently to Imposed Drought and Shade. Trees 2021, 35, 53–67. [Google Scholar] [CrossRef]
  41. Schuldt, B.; Knutzen, F.; Delzon, S.; Jansen, S.; Müller-Haubold, H.; Burlett, R.; Clough, Y.; Leuschner, C. How Adaptable Is the Hydraulic System of European Beech in the Face of Climate Change-related Precipitation Reduction? New Phytol. 2016, 210, 443–458. [Google Scholar] [CrossRef] [PubMed]
  42. Sommer, R.J. Phenotypic Plasticity: From Theory and Genetics to Current and Future Challenges. Genetics 2020, 215, 1–13. [Google Scholar] [CrossRef] [PubMed]
  43. Horváth, A.; Mátyás, C. The Decline of Vitality Caused by Increasing Drought in a Beech Provenance Trial Predicted by Juvenile Growth. SEEFOR 2016, 7, 21–28. [Google Scholar] [CrossRef]
  44. Sáenz-Romero, C.; Kremer, A.; Nagy, L.; Újvári-Jármay, É.; Ducousso, A.; Kóczán-Horváth, A.; Hansen, J.K.; Mátyás, C. Common Garden Comparisons Confirm Inherited Differences in Sensitivity to Climate Change between Forest Tree Species. PeerJ 2019, 7, e6213. [Google Scholar] [CrossRef] [PubMed]
  45. Robson, T.M.; Garzón, M.B. Phenotypic Trait Variation Measured on European Genetic Trials of Fagus Sylvatica L. Sci. Data 2018, 5, 180149. [Google Scholar] [CrossRef] [PubMed]
  46. Fick, S.E.; Hijmans, R.J. WorldClim 2: New 1-km Spatial Resolution Climate Surfaces for Global Land Areas. Int. J. Climatol. 2017, 37, 4302–4315. [Google Scholar] [CrossRef]
  47. Ziehn, T.; Chamberlain, M.; Lenton, A.; Law, R.; Bodman, R.; Dix, M.; Wang, Y.; Dobrohotoff, P.; Srbinovsky, J.; Stevens, L.; et al. CSIRO ACCESS-ESM1.5 Model Output Prepared for CMIP6 ScenarioMIP Ssp245. 2019. Available online: https://github.com/WCRP-CMIP/CMIP6_CVs (accessed on 19 December 2022).
  48. Zhang, Q.B.; Hebrda, J.H.; Zhang, Q.J.; Alfaro, R.I. Modeling Tree-Ring Growth Responses to Climatic Variables Using Artificial Neural Networks. For. Sci. 2000, 46, 229–239. [Google Scholar] [CrossRef]
  49. Khaki, S.; Wang, L.; Archontoulis, S.V. A CNN-RNN Framework for Crop Yield Prediction. Front. Plant Sci. 2020, 10, 1750. [Google Scholar] [CrossRef]
  50. Hornik, K. Approximation Capabilities of Multilayer Feedforward Networks. Neural Netw. 1991, 4, 251–257. [Google Scholar] [CrossRef]
  51. Montesinos López, O.A.; Montesinos López, A.; Crossa, J. Fundamentals of Artificial Neural Networks and Deep Learning. In Multivariate Statistical Machine Learning Methods for Genomic Prediction; Springer International Publishing: Cham, Switzerland, 2022; pp. 379–425. ISBN 978-3-030-89009-4. [Google Scholar]
  52. Pretzsch, H.; Hilmers, T.; Biber, P.; Avdagić, A.; Binder, F.; Bončina, A.; Bosela, M.; Dobor, L.; Forrester, D.I.; Lévesque, M.; et al. Evidence of Elevation-Specific Growth Changes of Spruce, Fir, and Beech in European Mixed Mountain Forests during the Last Three Centuries. Can. J. For. Res. 2020, 50, 689–703. [Google Scholar] [CrossRef]
  53. Hlásny, T.; Trombik, J.; Dobor, L.; Barcza, Z.; Barka, I. Future Climate of the Carpathians: Climate Change Hot-Spots and Implications for Ecosystems. Reg. Environ. Chang. 2016, 16, 1495–1506. [Google Scholar] [CrossRef]
  54. Čortan, D.; Nonić, M.; Šijačić-Nikolić, M. Phenotypic Plasticity of European Beech from International Provenance Trial in Serbia. In Forests of Southeast Europe Under a Changing Climate; Šijačić-Nikolić, M., Milovanović, J., Nonić, M., Eds.; Advances in Global Change Research; Springer International Publishing: Cham, Switzerland, 2019; Volume 65, pp. 333–351. ISBN 978-3-319-95266-6. [Google Scholar]
  55. Mátyás, C.; Berki, I.; Bidló, A.; Csóka, G.; Czimber, K.; Führer, E.; Gálos, B.; Gribovszki, Z.; Illés, G.; Hirka, A.; et al. Sustainability of Forest Cover under Climate Change on the Temperate-Continental Xeric Limits. Forests 2018, 9, 489. [Google Scholar] [CrossRef]
  56. Petkova, K.; Molle, E.; Konnert, M.; Knutzen, F. Comparing German and Bulgarian provenances of European beech (Fagus sylvatica L.) regarding survival, growth and ecodistance. Silva Balc. 2019, 2, 27–48. [Google Scholar] [CrossRef]
  57. Petrík, P.; Petek, A.; Konôpková, A.; Bosela, M.; Fleischer, P.; Frýdl, J.; Kurjak, D. Stomatal and Leaf Morphology Response of European Beech (Fagus sylvatica L.) Provenances Transferred to Contrasting Climatic Conditions. Forests 2020, 11, 1359. [Google Scholar] [CrossRef]
  58. Visi-Rajczi, E.; Hofmann, T.; Albert, L.; Mátyás, C. Tracing the Acclimation of European Beech (Fagus Sylvatica L.) Populations to Climatic Stress by Analyzing the Antioxidant System. iForest 2021, 14, 95–103. [Google Scholar] [CrossRef]
  59. Železnik, P.; Westergren, M.; Božič, G.; Eler, K.; Bajc, M.; Helmisaari, H.-S.; Horvath, A.; Kraigher, H. Root Growth Dynamics of Three Beech (Fagus sylvatica L.) Provenances. For. Ecol. Manag. 2019, 431, 35–43. [Google Scholar] [CrossRef]
  60. Bert, D.; Lebourgeois, F.; Ponton, S.; Musch, B.; Ducousso, A. Which Oak Provenances for the 22nd Century in Western Europe? Dendroclimatology in Common Gardens. PLoS ONE 2020, 15, e0234583. [Google Scholar] [CrossRef]
  61. Gömöry, D.; Longauer, R.; Hlásny, T.; Pacalaj, M.; Strmeň, S.; Krajmerová, D. Adaptation to Common Optimum in Different Populations of Norway Spruce (Picea abies Karst.). Eur. J. For. Res. 2012, 131, 401–411. [Google Scholar] [CrossRef]
  62. Jamnická, G.; Fleischer, P., Jr.; Konôpková, A.; Pšidová, E.; Kučerová, J.; Kurjak, D.; Živčák, M.; Ditmarová, L. Norway Spruce (Picea abies L.) Provenances Use Different Physiological Strategies to Cope with Water Deficit. Forests 2019, 10, 651. [Google Scholar] [CrossRef]
  63. Mátyás, C.; Bozic, G.; Gömöry, D.; Ivankovic, M.; Rasztovits, E. Juvenile Growth Response of European Beech (Fagus Sylvatica L.) to Sudden Change of Climatic Environment in SE European Trials. iForest 2009, 2, 213–220. [Google Scholar] [CrossRef]
  64. Martinez-Vilalta, J.; Anderegg, W.R.L.; Sapes, G.; Sala, A. Greater Focus on Water Pools May Improve Our Ability to Understand and Anticipate Drought-induced Mortality in Plants. New Phytol. 2019, 223, 22–32. [Google Scholar] [CrossRef] [PubMed]
  65. Lévesque, M.; Walthert, L.; Weber, P. Soil Nutrients Influence Growth Response of Temperate Tree Species to Drought. J. Ecol. 2016, 104, 377–387. [Google Scholar] [CrossRef]
  66. Kreuzwieser, J.; Gessler, A. Global Climate Change and Tree Nutrition: Influence of Water Availability. Tree Physiol. 2010, 30, 1221–1234. [Google Scholar] [CrossRef] [PubMed]
  67. Gustafson, E.; Miranda, B.; Sturtevant, B. Can Future CO2 Concentrations Mitigate the Negative Effects of High Temperature and Longer Droughts on Forest Growth? Forests 2018, 9, 664. [Google Scholar] [CrossRef]
  68. De Kauwe, M.G.; Medlyn, B.E.; Tissue, D.T. To What Extent Can Rising CO2 Ameliorate Plant Drought Stress? New Phytol. 2021, 231, 2118–2124. [Google Scholar] [CrossRef] [PubMed]
  69. Bussotti, F.; Pollastrini, M. Traditional and Novel Indicators of Climate Change Impacts on European Forest Trees. Forests 2017, 8, 137. [Google Scholar] [CrossRef]
  70. Müller, M.; Kempen, T.; Finkeldey, R.; Gailing, O. Low Population Differentiation but High Phenotypic Plasticity of European Beech in Germany. Forests 2020, 11, 1354. [Google Scholar] [CrossRef]
  71. Harter, D.E.V.; Nagy, L.; Backhaus, S.; Beierkuhnlein, C.; Fussi, B.; Huber, G.; Jentsch, A.; Konnert, M.; Thiel, D.; Kreyling, J. A Comparison of Genetic Diversity and Phenotypic Plasticity among European Beech (Fagus sylvatica L.) Populations from Bulgaria and Germany under Drought and Temperature Manipulation. Int. J. Plant Sci. 2015, 176, 232–244. [Google Scholar] [CrossRef]
  72. Frank, A.; Pluess, A.R.; Howe, G.T.; Sperisen, C.; Heiri, C. Quantitative Genetic Differentiation and Phenotypic Plasticity of European Beech in a Heterogeneous Landscape: Indications for Past Climate Adaptation. Perspect. Plant Ecol. Evol. Syst. 2017, 26, 1–13. [Google Scholar] [CrossRef]
  73. Kramer, K.; Ducousso, A.; Gömöry, D.; Hansen, J.K.; Ionita, L.; Liesebach, M.; Lorenţ, A.; Schüler, S.; Sulkowska, M.; de Vries, S.; et al. Chilling and Forcing Requirements for Foliage Bud Burst of European Beech (Fagus sylvatica L.) Differ between Provenances and Are Phenotypically Plastic. Agric. For. Meteorol. 2017, 2017, 172–181. [Google Scholar] [CrossRef]
  74. Wang, H.; Lin, S.; Dai, J.; Ge, Q. Modeling the Effect of Adaptation to Future Climate Change on Spring Phenological Trend of European Beech (Fagus sylvatica L.). Sci. Total Environ. 2022, 846, 157540. [Google Scholar] [CrossRef] [PubMed]
  75. Stojnić, S.; Orlović, S.; Miljković, D.; Galić, Z.; Kebert, M.; von Wuehlisch, G. Provenance Plasticity of European Beech Leaf Traits under Differing Environmental Conditions at Two Serbian Common Garden Sites. Eur. J. For. Res. 2015, 134, 1109–1125. [Google Scholar] [CrossRef]
  76. Vastag, E.; Kovacevic, B.; Orlovic, S.; Kesic, L.; Bojovic, M.; Stojnic, S. Leaf Stomatal Traits Variation within and among Fourteen European Beech (Fagus sylvatica L.) Provenances. Genetika 2019, 51, 937–959. [Google Scholar] [CrossRef]
  77. Stojnić, S.; Viscosi, V.; Marković, M.; Ivanković, M.; Orlović, S.; Tognetti, R.; Cocozza, C.; Vasić, V.; Loy, A. Spatial Patterns of Leaf Shape Variation in European Beech (Fagus sylvatica L.) Provenances. Trees 2022, 36, 497–511. [Google Scholar] [CrossRef]
  78. Stojnić, S.; Suchocka, M.; Benito-Garzón, M.; Torres-Ruiz, J.M.; Cochard, H.; Bolte, A.; Cocozza, C.; Cvjetković, B.; de Luis, M.; Martinez-Vilalta, J.; et al. Variation in Xylem Vulnerability to Embolism in European Beech from Geographically Marginal Populations. Tree Physiol. 2018, 38, 173–185. [Google Scholar] [CrossRef]
  79. Vicente, E.; Didion-Gency, M.; Morcillo, L.; Morin, X.; Vilagrosa, A.; Grossiord, C. Aridity and Cold Temperatures Drive Divergent Adjustments of European Beech Xylem Anatomy, Hydraulics and Leaf Physiological Traits. Tree Physiol. 2022, 42, 1720–1735. [Google Scholar] [CrossRef]
  80. Weithmann, G.; Link, R.M.; Banzragch, B.-E.; Würzberg, L.; Leuschner, C.; Schuldt, B. Soil Water Availability and Branch Age Explain Variability in Xylem Safety of European Beech in Central Europe. Oecologia 2022, 198, 629–644. [Google Scholar] [CrossRef]
  81. Miranda, J.C.; Calderaro, C.; Cocozza, C.; Lasserre, B.; Tognetti, R.; von Arx, G. Wood Anatomical Responses of European Beech to Elevation, Land Use Change, and Climate Variability in the Central Apennines, Italy. Front. Plant Sci. 2022, 13, 855741. [Google Scholar] [CrossRef]
  82. Pšidová, E.; Živčák, M.; Stojnić, S.; Orlović, S.; Gömöry, D.; Kučerová, J.; Ditmarová, Ľ.; Střelcová, K.; Brestič, M.; Kalaji, H.M. Altitude of Origin Influences the Responses of PSII Photochemistry to Heat Waves in European Beech (Fagus sylvatica L.). Environ. Exp. Bot. 2018, 152, 97–106. [Google Scholar] [CrossRef]
  83. Dziedek, C.; Fichtner, A.; Calvo, L.; Marcos, E.; Jansen, K.; Kunz, M.; Walmsley, D.; Von Oheimb, G.; Härdtle, W. Phenotypic Plasticity Explains Response Patterns of European Beech (Fagus sylvatica L.) Saplings to Nitrogen Fertilization and Drought Events. Forests 2017, 8, 91. [Google Scholar] [CrossRef]
  84. Weithmann, G.; Schuldt, B.; Link, R.M.; Heil, D.; Hoeber, S.; John, H.; Müller-Haubold, H.; Schüller, L.-M.; Schumann, K.; Leuschner, C. Leaf Trait Modification in European Beech Trees in Response to Climatic and Edaphic Drought. Plant Biol. 2021, 24, 1272–1286. [Google Scholar] [CrossRef] [PubMed]
  85. Dulamsuren, C.; Hauck, M.; Kopp, G.; Ruff, M.; Leuschner, C. European Beech Responds to Climate Change with Growth Decline at Lower, and Growth Increase at Higher Elevations in the Center of Its Distribution Range (SW Germany). Trees 2017, 31, 673–686. [Google Scholar] [CrossRef]
  86. Muffler, L.; Schmeddes, J.; Weigel, R.; Barbeta, A.; Beil, I.; Bolte, A.; Buhk, C.; Holm, S.; Klein, G.; Klisz, M.; et al. High Plasticity in Germination and Establishment Success in the Dominant Forest Tree Fagus sylvatica across Europe. Global Ecol. Biogeogr. 2021, 30, 1583–1596. [Google Scholar] [CrossRef]
  87. Piovesan, G.; Biondi, F.; Filippo, A.D.; Alessandrini, A.; Maugeri, M. Drought-Driven Growth Reduction in Old Beech (Fagus Sylvatica L.) Forests of the Central Apennines, Italy: Drought-driven growth decline in beech forests. Glob. Change Biol. 2008, 14, 1265–1281. [Google Scholar] [CrossRef]
  88. Knutzen, F.; Dulamsuren, C.; Meier, I.C.; Leuschner, C. Recent Climate Warming-Related Growth Decline Impairs European Beech in the Center of Its Distribution Range. Ecosystems 2017, 20, 1494–1511. [Google Scholar] [CrossRef]
  89. Aertsen, W.; Janssen, E.; Kint, V.; Bontemps, J.-D.; Van Orshoven, J.; Muys, B. Long-Term Growth Changes of Common Beech (Fagus Sylvatica L.) Are Less Pronounced on Highly Productive Sites. For. Ecol. Manag. 2014, 312, 252–259. [Google Scholar] [CrossRef]
  90. Zimmermann, J.; Hauck, M.; Dulamsuren, C.; Leuschner, C. Climate Warming-Related Growth Decline Affects Fagus Sylvatica, But Not Other Broad-Leaved Tree Species in Central European Mixed Forests. Ecosystems 2015, 18, 560–572. [Google Scholar] [CrossRef]
  91. Vacek, Z.; Prokůpková, A.; Vacek, S.; Bulušek, D.; Šimůnek, V.; Hájek, V.; Králíček, I. Mixed vs. Monospecific Mountain Forests in Response to Climate Change: Structural and Growth Perspectives of Norway Spruce and European Beech. For. Ecol. Manag. 2021, 488, 119019. [Google Scholar] [CrossRef]
  92. Sedmáková, D.; Sedmák, R.; Bosela, M.; Ježík, M.; Blaženec, M.; Hlásny, T.; Marušák, R. Growth-Climate Responses Indicate Shifts in the Competitive Ability of European Beech and Norway Spruce under Recent Climate Warming in East-Central Europe. Dendrochronologia 2019, 54, 37–48. [Google Scholar] [CrossRef]
  93. Prislan, P.; Gričar, J.; Čufar, K.; de Luis, M.; Merela, M.; Rossi, S. Growing Season and Radial Growth Predicted for Fagus Sylvatica under Climate Change. Clim. Change 2019, 153, 181–197. [Google Scholar] [CrossRef]
  94. Richardson, A.D.; Andy Black, T.; Ciais, P.; Delbart, N.; Friedl, M.A.; Gobron, N.; Hollinger, D.Y.; Kutsch, W.L.; Longdoz, B.; Luyssaert, S.; et al. Influence of Spring and Autumn Phenological Transitions on Forest Ecosystem Productivity. Phil. Trans. R. Soc. B 2010, 365, 3227–3246. [Google Scholar] [CrossRef] [PubMed]
  95. Schieber, B.; Kubov, M.; Janík, R. Effects of Climate Warming on Vegetative Phenology of the Common Beech Fagus Sylvatica in a Submontane Forest of the Western Carpathians: Two-Decade Analysis. Pol. J. Ecol. 2017, 65, 339–351. [Google Scholar] [CrossRef]
  96. Mura, C.; Buttò, V.; Silvestro, R.; Deslauriers, A.; Charrier, G.; Raymond, P.; Rossi, S. The Early Bud Gets the Cold: Diverging Spring Phenology Drives Exposure to Late Frost in a Picea mariana [(Mill.) BSP] Common Garden. Physiol. Plant. 2022, 174, e13798. [Google Scholar] [CrossRef] [PubMed]
  97. Vitali, A.; Tonelli, E.; Malandra, F.; Camarero, J.J.; Colangelo, M.; Nolè, A.; Ripullone, F.; Carrer, M.; Urbinati, C. Combining Dendroecology and Remote Sensing to Assess How Late Spring Frosts Affect European Beech Forests. EGU Gen. Assem. 2022, EGU22-5514. [Google Scholar] [CrossRef]
Figure 1. Geographical locations of European beech provenances (a) and common garden sites (b) used in this study.
Figure 1. Geographical locations of European beech provenances (a) and common garden sites (b) used in this study.
Forests 14 00026 g001
Figure 2. Average annual precipitation and average annual temperature at the sites of provenance origin (1961–1990, (a)) and at the common garden sites (1995–2002/1998–2005, (b)).
Figure 2. Average annual precipitation and average annual temperature at the sites of provenance origin (1961–1990, (a)) and at the common garden sites (1995–2002/1998–2005, (b)).
Forests 14 00026 g002
Figure 3. The comparison of observed and predicted values by the neural network model (a) and residual scatter plot (b).
Figure 3. The comparison of observed and predicted values by the neural network model (a) and residual scatter plot (b).
Forests 14 00026 g003
Figure 4. Sensitivity analysis of standardised variables within the neural network model (a), Bio13 = precipitation of wettest month (b), Bio8 = mean temperature of wettest quarter (c,d), Bio12 = annual precipitation (e), Bio14 = precipitation of driest month (f), Bio18 = precipitation of warmest quarter (g,h). Bio variables with ED subscript represent the ecodistance, which is the difference between the climate at the site of origin and the climate of new sites.
Figure 4. Sensitivity analysis of standardised variables within the neural network model (a), Bio13 = precipitation of wettest month (b), Bio8 = mean temperature of wettest quarter (c,d), Bio12 = annual precipitation (e), Bio14 = precipitation of driest month (f), Bio18 = precipitation of warmest quarter (g,h). Bio variables with ED subscript represent the ecodistance, which is the difference between the climate at the site of origin and the climate of new sites.
Forests 14 00026 g004
Figure 5. Maps of simulated height growth for four contrasting provenance groups during seven years under 2040–2060 climate derived from the SSP245/RCP4.5 climate scenario (a—hotdry origin, b—hotwet origin, c—colddry origin, d—coldwet origin). The height is simulated for trees that are approximately 10 years old at the end of the simulation. The state on the left side is the Czech Republic, and the state on the right side is Slovakia.
Figure 5. Maps of simulated height growth for four contrasting provenance groups during seven years under 2040–2060 climate derived from the SSP245/RCP4.5 climate scenario (a—hotdry origin, b—hotwet origin, c—colddry origin, d—coldwet origin). The height is simulated for trees that are approximately 10 years old at the end of the simulation. The state on the left side is the Czech Republic, and the state on the right side is Slovakia.
Forests 14 00026 g005
Figure 6. Mean and 95% confidence intervals of simulated height growth (2040–2060), for 20 provenances, from 4 provenance groups with contrasting original climate. Boxplots on provenance group level with lowercase letters showing results of Tukey HSD post-hoc test. AU—Austria, CR—Croatia, CZ—Czech Republic, ESP—Spain, FR—France, GER—Germany, IT—Italy, PL—Poland, RO—Romania, UA—Ukraine.
Figure 6. Mean and 95% confidence intervals of simulated height growth (2040–2060), for 20 provenances, from 4 provenance groups with contrasting original climate. Boxplots on provenance group level with lowercase letters showing results of Tukey HSD post-hoc test. AU—Austria, CR—Croatia, CZ—Czech Republic, ESP—Spain, FR—France, GER—Germany, IT—Italy, PL—Poland, RO—Romania, UA—Ukraine.
Forests 14 00026 g006
Table 1. Results of two-way ANOVA for height variability among provenance populations and common garden sites.
Table 1. Results of two-way ANOVA for height variability among provenance populations and common garden sites.
FactorDfSumSqMeanSqFp
Provenance13071,289,815548,383109.66<0.001
Site13188,718,15014,516,7812902.8<0.001
Provenance × Site40922,573,97955,19311.04<0.001
Residuals45,735228,715,6825001
Table 2. Climatic boundary conditions of the neural network model based on minimal and maximal values from the original dataset used for development of the model.
Table 2. Climatic boundary conditions of the neural network model based on minimal and maximal values from the original dataset used for development of the model.
Bio8Bio14Bio18Bio8EDBio12EDBio13EDBio18ED
Min0.11278−16.4−928−122−342
Max19.988533171261138319
Unit°Cmmmm°CmmmmMm
Table 3. Annual mean temperature (T) and annual precipitation sum (P) for original location of provenance groups used for simulation of growth, divided into contrasting climate groups. Full names of provenances are listed in Supplementary Table S2.
Table 3. Annual mean temperature (T) and annual precipitation sum (P) for original location of provenance groups used for simulation of growth, divided into contrasting climate groups. Full names of provenances are listed in Supplementary Table S2.
PVT (°C)P (mm)PVT (°C)P (mm)
Hot–DryHot–Wet
FR0210.6682ESP029.8967
ESP0510.5633FR239.8975
GER478.8556IT3712.31251
IT7813.8595IT10810.91080
IT808.2762CR13914.41271
Cold–DryCold–Wet
PL385.9624AU352.41524
CZ484.4769AU3641184
GER855.1709CZ514.61124
PL1174.9757AU393.81067
RO1556.5583UA1414.9943
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Petrík, P.; Grote, R.; Gömöry, D.; Kurjak, D.; Petek-Petrik, A.; Lamarque, L.J.; Sliacka Konôpková, A.; Mukarram, M.; Debta, H.; Fleischer, P., Jr. The Role of Provenance for the Projected Growth of Juvenile European Beech under Climate Change. Forests 2023, 14, 26. https://doi.org/10.3390/f14010026

AMA Style

Petrík P, Grote R, Gömöry D, Kurjak D, Petek-Petrik A, Lamarque LJ, Sliacka Konôpková A, Mukarram M, Debta H, Fleischer P Jr. The Role of Provenance for the Projected Growth of Juvenile European Beech under Climate Change. Forests. 2023; 14(1):26. https://doi.org/10.3390/f14010026

Chicago/Turabian Style

Petrík, Peter, Rüdiger Grote, Dušan Gömöry, Daniel Kurjak, Anja Petek-Petrik, Laurent J. Lamarque, Alena Sliacka Konôpková, Mohammad Mukarram, Harish Debta, and Peter Fleischer Jr. 2023. "The Role of Provenance for the Projected Growth of Juvenile European Beech under Climate Change" Forests 14, no. 1: 26. https://doi.org/10.3390/f14010026

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop