Physiological tolerance to frost and drought explains range limits of 35 European tree species
Data files
Nov 22, 2024 version files 1.52 MB
-
era5_svwcperc05d_h1_1984-2021_inv.grib
302.70 KB
-
era5_svwcperc05d_h2_1984-2021_inv.grib
302.70 KB
-
era5_svwcperc05d_h3_1984-2021_inv.grib
302.70 KB
-
era5_svwcperc05d_h4_1984-2021_inv.grib
302.70 KB
-
era5_t2min05_1984-2021.grib
295.23 KB
-
README.md
4.67 KB
-
Trait_ecog_filtered_dryad.csv
6.46 KB
Abstract
Models are key to evaluating how climate change threatens European forests and tree species distributions. However, current species distribution models struggle to integrate ecophysiological processes. Mechanistic models are complex and have high parameter requirements. Some correlative species distribution models have tried to include traits but so far have failed to directly connect to ecophysiological processes. Here, we propose an intermediate strategy where species distributions are based on their safety margins which represent their proximity to their physiological thresholds.
We derived frost and drought safety margins for 38 European tree species as the difference between physiological tolerance traits and local maximum stress. We used LT50 and P50 as tolerance traits for frost and drought, respectively, and local minimum temperature and minimum soil water potential as maximum stress. We integrated these safety margins into a species distribution model, which tests if the probability of species presence declines rapidly when the safety margin reaches zero, when physiological stress exceeds the species' tolerance traits.
Our results showed that 35 of the 38 studied species had their distribution explained by one or both safety margins. We demonstrated that safety-margins-based model can be efficiently transferred to species for which occurrence data are not available.
The probability of presence dropped dramatically when the frost safety margin reached zero, whereas it was less sensitive to the drought safety margin. This differential sensitivity may be due to the more complex regulation of drought stress, especially as water is a shared resource, whereas frost is not. However, we did not find a clear effect of local competition hierarchy on the response to safety margin, but competition needs further analysis.
Our analysis provides a new approach to linking species distributions to their physiological limits and shows that, in Europe, frost and drought safety margins are important determinants of species distributions.
README: Physiological tolerance to frost and drought explains range limits of 35 European tree species
https://doi.org/10.5061/dryad.547d7wmh1
This repository provides the data and code required to compute safety margins of 38 tree species across Europe. The code enables us to fit distribution models based on these safety margins, and to compute performance metrics of models.
Description of the data
Trait_ecog_filtered_dryad.csv
gathers stress tolerance traits used in the analysis, for the 38 European species. This database results from the aggregation of data from a literature review. "class", "order", "family", "genus" and "taxa" are phylogenetic information. "lt50" and "lt50sd" are the raw mean, and standard deviation of frost resistance computed from the literature review. "lt50source" indicates on which anatomical structure the frost resistance was measured. P12, P50, and P88 are the measures of tree resistance to cavitation, and P50sd is its standard deviation. "px" is the final value chosen among P50 and P88 for the species. "pxsource" indicates from which database the trait was extracted. "slope" and "slopemsp" are measures of the slopes of the vulnerability curves.
era5_svwcperc05d_X_1984-2021_inv
files (with X being h1, h2, h3 or h4), are the minimum soil water content computed over the four soil horizons across Europe. Each layer was derived from monthly data obtained from ERA5-land climate reanalysis accessible at https://cds.climate.copernicus.eu/datasets/reanalysis-era5-land?tab=download
era5_t2min05_1984-2021.grib
is the minimum temperature. It is also derived from ERA5-land reanalysis.
Access to other required data
Other data needs to be downloaded to run the analysis.
Hydraulic parameters
Hydraulic parameters of soil layers can be downloaded from EU_SoilHydroGrids database.
Rooting depth
Soil rooting depth can be downloaded from the European Soil DataBase
Mean climate variables
Mean annual temperature (bio1), mean annual precipitation (bio12), growing degree days (gdd5) and potential evapotranspiration (pet_penman) were download from CHELSA over the period 1981-2010.
Species occurrence
We used EuForest data from Mauri et al. (2017) to get occurrence data of our 38 tree species.
Species distribution expert maps
The analysis requires maps of species distribution from expert-knowledge. These were download from the EuForgen project at this path.
Biomes map
Map of global biomes can be downloaded from WWF website.
Description of the code
The main worflow is written with the targets package, and all you need to do is run the tar_make() function in R.
The workflow functions are written in separate R files. Finally, a figure file is used to load objects and create figures. The whole workflow takes quite a long time to complete.
.stan files are the files of the different models run during the analysis.
Data folder architecture
To run the whole code, we suggest users to reproduce the same arborescence as we used in our analysis, which implies a data folder with the following organisation:
- ERA5-land
-- daily
---era5_svwcperc05d_h1_1984-2021_inv
---era5_svwcperc05d_h2_1984-2021_inv
---era5_svwcperc05d_h3_1984-2021_inv
---era5_svwcperc05d_h4_1984-2021_inv
--t2m
---era5_t2min05_1984-2021.grib
- CHELSA (files to download as describe above)
-- Bio1 file
-- Bio12 file
-- Gdd5 file
-- Pet_penman file
- species trait
-- base_traits_P50_LTx.xlsx
- STU_EU_Layers (files to download as describe above in Rooting depth)
-- soil rooting depth file
- EU_SoilHydroGrids_1km (files to download as describe above in Hydraulic parameters)
-- all hydraulic parameters files
- WWF (files to download as describe above in Biomes map)
-- official
--- all files downloaded
- chronologica_maps_dataset (files to download as describe above in Species distribution expert maps)
-- all file downloaded
- EUForestsMauri (files to download as describe above in Species occurrence )
--all files downloaded