Climate change accelerates ecosystem restoration in the mountain forests of Central Europe
Data files
Oct 05, 2023 version files 221.93 MB
Abstract
- Restoring degraded forest ecosystems is an important element in the ongoing challenge to sustain the integrity and functioning of the biosphere. However, the evaluation of restoration success is hampered by long lead times of management measures in forests. Moreover, forest change is accelerating in the absence of management because of ongoing climate change. Yet, because a counterfactual is frequently missing, it remains unclear whether restoration measures are aided or impeded by climate change.
- Here, we analyzed the pace and success of forest restoration under climate change, combining field data and simulation modelling. We focused on the management zone of Berchtesgaden National Park (BGNP), Germany, where restoration aims to restore homogeneous Norway spruce (Picea abies) forests to structurally diverse mixed mountain forests. We evaluated three alternative restoration strategies: Two active strategies focused on planting the currently underrepresented silver fir (Abies alba) and European beech (Fagus sylvatica) but differing in the creation of gap-cuts, and a third passive restoration strategy without interventions. Strategies were simulated with the forest landscape model iLand from 2020 to 2100 under different climate scenarios (historic, RCP 2.6, 4.5, and 8.5).
- The forests of BGNP developed into structurally diverse and mixed forests under all evaluated management strategies, and differences between active and passive restoration were generally small. While restoration goals for forest structure were largely met by 2100, forest composition remained far from target in all strategies. Climate change aided restoration by significantly increasing the prevalence of silver fir and European beech (+104.2 % to +258.6 %). Field data on short-term restoration effects were in line with simulated long-term trajectories.
- Synthesis and applications. We here show that forest restoration efforts in Central European mountain forests will likely be accelerated by climate change. Nonetheless, the slow pace of restoration underscores the need for taking action. Our study highlights that active restoration measures such as tree planting can bring the system closer to restoration targets. However, it also demonstrates that passive restoration (no intervention) is a viable option for management, highlighting the need to evaluate restoration measures against the counterfactual of a no-intervention strategy.
README: Climate change accelerates ecosystem restoration in the mountain forests of Central Europe
https://doi.org/10.5061/dryad.v15dv422r
The data from "fieldData" was collected in field experiments to measure the short-term restoration effects.
The data from "simulationData" was simulated using the forst simulation model iLand (https://iland-model.org) to illuminate the long-term restoration effects.
Description of the data and file structure
fieldData: for each sampled plot, the tree species found in the regeneration layer were recorded
column | description | |
---|---|---|
plot | plot identifier | |
treatment | "No Intervention": control sample of passive restoration; "With Intervention": active restoration | |
species | observed tree species; "Bah" = Acer pseudoplatanus; "Bi" = Betula pendula; "Bu" = Fagus sylvatica; "BUlm" = Ulmus glabra; "Eibe" = Taxus baccata; "Erle" = Alnus glutinosa; "Esche" = Fraxinus excelsior; "Fi" = Picea abies; "Lä" = Larix decidua; "Mehlbeere" = Sorbus aria; "Salweide" = Salix caprea; "Hasel" = Corylus avellana; "Ta" = Abies alba; "Vogelb." = Sorbus aucuparia; "Zitterpappel" = Populus tremula | |
scenario | data collection stratum; "reactive": regeneration in windthrow patches with and without planting; "proactive": regeneration in mature Norway spruce stands treated with gap cuts relative to those without intervention |
simulationData: for each simulation multiple outputs about forest composition and structure as well as simulated management interventions were generated
Since the outputs about forest composition and structure for the 120 simulation runs are each ~200-250 MB in size, only one is uploaded as an example. The other files can be requested from the corresponding author, Christina Dollinger.
1. output_AG (.csv\, n=40)
column | description |
---|---|
year | simulation year (1-80) |
id | ID of the Ressource Unit where an artifical gap had been created in the specific simulation year |
nTrees | Number of trees cut in the 50x40 m artificial gap |
2. output_DG (.csv\, n=80)
column | description |
---|---|
year | simulation year (1-80) |
threshold | fixed threshold (0.2); minimum proportion of trees that needed to killed by bark beetles or wind in a continuous patch of 1,600 m² for it to be considered a disturbance patch |
id | ID of the Ressource Unit where a disturbance patch had been identified in the specific simulation year |
3. output (.sqlite\, n=120\, only 1 uploaded here)
table | description |
---|---|
barkbeetle | https://iland-model.org/barkbeetle+module#BarkBeetle_module_output |
dynamicstand | https://iland-model.org/Outputs?highlight=basal+area#dynamic_stand_output_by_species_RU |
landscape | https://iland-model.org/Outputs?highlight=basal+area#Landscape_aggregates_per_species |
landscape_removed | https://iland-model.org/Outputs?highlight=basal+area#Aggregates_of_removed_trees_due_to_death_harvest_and_disturbances_per_species |
runinfo | timestamp when the simulation was started, and iLand version information |
stand | https://iland-model.org/Outputs?highlight=basal+area#Stand_by_species_RU |
wind | https://iland-model.org/wind+module#Wind_disturbance_module_output |
Code/Software
Data preparation as well as all analyses were done using the R project for statistical computing version 4.0.5 (R Core Team, 2021).