Data from: Shade tolerance controls the spectrum of crown sizes and its response to local competition across European and North American tree species: Implications for light interception strategies
Data files
May 07, 2025 version files 24.76 MB
-
README.md
25.87 KB
-
tree_crown_allometry.zip
24.73 MB
Abstract
The size of a tree and its crown are key drivers of its ability to intercept light. These characteristics are also constrained by biomechanical and hydraulic limitations, leading to tremendous variations in crown size and tree height between species. To date, there is no consensus on how covariation between crown characteristics (i.e., tree height, crown diameter and relative crown depth) control the spectrum of crown sizes, nor on the species characteristics (i.e., species climate niche or functional traits) that govern their variation. Using species-specific allometric relationships accounting for local competition data on individual crown characteristics of tree species across Europe and North America, we showed that the spectrum of tree crown size is mainly driven by two independent axes: one correlating tree height with relative crown depth, and another linking crown diameter to crown volume. While both angiosperms and gymnosperms exhibited similar patterns of covariation for these crown characteristics, distinct ecological strategies emerged for maximum tree height. We also highlighted that shade-tolerant species were characterized by wider and deeper crowns and showed reduced sensitivity to local competition in relative crown depth, indicating enhanced competitive resilience. Finally, simulations of light interception in virtual stands performed using a ray tracing algorithm highlighted the importance of different crown characteristics in various competitive contexts. Tree height was the primary determinant of light interception in low competitive contexts, common in early successional stages, whereas relative crown depth and crown diameter played a more significant role in high-competitive contexts typical of late succession. These findings suggest that shade-tolerant species, with their deeper crowns, are better equipped to intercept light under competition and ultimately emphasize the ability of shade tolerance to improve the realism of forest dynamic models as it can be used as an indicator of crown size and its response to competition.
Dataset DOI: 10.5061/dryad.612jm64g4
Description of the data and file structure
Overall data description
This repository provides the data required to replicate the analysis of the study entitled "Shade tolerance controls the spectrum of crown sizes and its response to local competition across European and North American tree species. Implications for light interception strategies*.*"
The code to implement all analyses is available in the following GitHub repository: https://github.com/LauraTouzot/Tree_crown_allometry_DECLIC_Project.git
The parameters of all fitted allometric equations for crown diameter, relative crown depth and tree height for all species used in the analysis are provided. In addition, the estimated light interception computed with SamsaraLight are provided along the functional traits used in the analyses.
Files and variables
File: tree_crown_allometry.zip
Description: data are organised into folders, with one folder corresponding to each studied crown characteristic (i.e., crown diameter, relative crown depth and tree height). Within each of these folders, the two competition conditions (i.e. with or without competition) are separated into sub-folders. Below is a detailed description of all the folders and the files they contain. Folders and subfolders are shown in bold, files in italics.
Crown diameter
- Competition
- crowndiameter_mean_parameters_competition.csv : average parameters obtained from 100 resamplings for each species studied and each of the parameters defined in the selected allometric power-law relationship linking crown diameter to tree diameter at breast height (i.e., weighted protocol effect and Basal Area of trees Larger than the focal tree as competition index); a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2, ci corresponds to parameter cj defined in equation 4; cd stands for crown diameter and mean and sd represent the mean of the parameter and the standard deviation obtained over the 100 resamplings, respectively
- diameter_comp_weighted_parameters_c1.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric power-law relationship linking crown diameter to tree diameter at breast height (equation 2) and with BAT (Total Basal Area) as competition index (equation 4) and a weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2, comp corresponds to parameter cj defined in equation 4
- diameter_comp_weighted_parameters_c2.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric power-law relationship linking crown diameter to tree diameter at breast height (equation 2) and with BAL as competition index (equation 4) and a weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2, comp corresponds to parameter cj defined in equation 4
- diameter_comp_no_weighted_parameters_c1.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric power-law relationship linking crown diameter to tree diameter at breast height (equation 2) and with BAT (Total Basal Area) as competition index (equation 4) and a no-weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2, comp corresponds to parameter cj defined in equation 4
- diameter_comp_no_weighted_parameters_c2.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric power-law relationship linking crown diameter to tree diameter at breast height (equation 2) and with BAL as competition index (equation 4) and a no-weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2, comp corresponds to parameter cj defined in equation 4
- diameter_comp_RMSE_AIC_results.csv: mean ( _mean_) and total ( _total_) RMSE values, and mean and total AIC values, obtained over all 100 resamplings for each studied species with either a weighted or no weighted protocol effect included in the power-law relationship (equation 2), and with either BAT (c1) or BAL (c2) included as the competition index (see equation 4). Note that in the case of AIC, only the weighted protocol effect in the models was considered based on the results obtained with RMSE.
- No competition
- crowndiameter_mean_parameters_nocompetition.csv: average parameters obtained from 100 resamplings for each species studied and each of the parameters defined in the selected allometric power-law relationship linking crown diameter to tree diameter at breast height (i.e., weighted protocol effect); a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2; cd stands for crown diameter and mean and sd represent the mean of the parameter and the standard deviation obtained over the 100 resamplings, respectively
- diameter_nocomp_no_weighted_parameters.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric power-law relationship linking crown diameter to tree diameter at breast height (equation 2) with a no-weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2
- diameter_nocomp_weighted_parameters.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric power-law relationship linking crown diameter to tree diameter at breast height (equation 2) with a weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2
- diameter_nocomp_RMSE_results.csv: mean ( _mean_) and total ( _total_) RMSE values obtained over all 100 resamplings for each studied species with either a weighted or no weighted protocol effect included in the power-law relationship (equation 2).
Relative crown depth
- Competition
- relativecrowndepth_mean_parameters_competition.csv : average parameters obtained from 100 resamplings for each species studied and each of the parameters defined in the selected allometric beta regression linking relative crown depth to tree diameter at breast height (i.e., weighted protocol effect and Basal Area of trees Larger than the focal tree as competition index); a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 3, ci corresponds to parameter cj defined in equation 4; rcd stands for relative crown depth and mean and sd represent the mean of the parameter and the standard deviation obtained over the 100 resamplings, respectively
- ratio_comp_weighted_parameters_c1.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric beta regression linking relative crown depth to tree diameter at breast height (equation 3) and with BAT as competition index (equation 4) and a weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 3, comp corresponds to parameter cj defined in equation 4
- ratio_compweighted_parameters_c2.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric beta regression linking relative crown depth to tree diameter at breast height (equation 3) and with BAL as competition index (equation 4) and a weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 3, comp corresponds to parameter cj defined in equation 4
- ratio_compno_weighted_parameters_c1.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric beta regression linking relative crown depth to tree diameter at breast height (equation 3) and with BAT as competition index (equation 4) and a no-weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 3, comp corresponds to parameter cj defined in equation 4
- ratio_compno_weighted_parameters_c2.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric beta regression linking relative crown depth to tree diameter at breast height (equation 3) and with BAL as competition index (equation 4) and a no-weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 3, comp corresponds to parameter cj defined in equation 4
- ratio_compRMSE_AIC_results.csv: mean ( _mean_) and total ( _total_) RMSE values, and mean and total AIC values, obtained over all 100 resamplings for each studied species with either a weighted or no weighted protocol effect included in the beta regression (equation 3), and with either BAT (c1) or BAL (c2) included as the competition index (see equation 4). Note that in the case of AIC, only the weighted protocol effect in the models was considered based on the results obtained with RMSE.
- No competition
- relativecrowndepth_mean_parameters_nocompetition.csv: average parameters obtained from 100 resamplings for each species studied and each of the parameters defined in the selected allometric beta regression linking relative crown depth to tree diameter at breast height (i.e., weighted protocol effect); a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 3; rcd stands for relative crown depth and mean and sd represent the mean of the parameter and the standard deviation obtained over the 100 resamplings, respectively
- ratio_nocomp_no_weighted_parameters.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric beta regression linking relative crown depth to tree diameter at breast height (equation 3) with a no-weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 3
- ratio_nocomp_weighted_parameters.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric beta regression linking relative crown depth to tree diameter at breast height (equation 3) with a weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2
- ratio_nocomp_RMSE_results.csv: mean ( _mean_) and total ( _total_) RMSE values obtained over all 100 resamplings for each studied species with either a weighted or no weighted protocol effect included in the beta regression (equation 3).
Tree height
- Competition
- treeheight_mean_parameters_competition.csv : average parameters obtained from 100 resamplings for each species studied and each of the parameters defined in the selected allometric power-law relationship linking tree height to tree diameter at breast height (i.e., weighted protocol effect and Basal Area of trees Larger than the focal tree as competition index); a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2, ci corresponds to parameter cj defined in equation 4; h stands for tree height and mean and sd represent the mean of the parameter and the standard deviation obtained over the 100 resamplings, respectively
- height_power_comp_weighted_parameters_c1.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric power-law relationship linking tree height to tree diameter at breast height (equation 2) and with BAT (Total Basal Area) as competition index (equation 4) and a weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2, comp corresponds to parameter cj defined in equation 4
- height_power_comp_weighted_parameters_c2.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric power-law relationship linking tree height to tree diameter at breast height (equation 2) and with BAL as competition index (equation 4) and a weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2, comp corresponds to parameter cj defined in equation 4
- height_power_comp_no_weighted_parameters_c1.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric power-law relationship linking tree height to tree diameter at breast height (equation 2) and with BAT (Total Basal Area) as competition index (equation 4) and a no-weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2, comp corresponds to parameter cj defined in equation 4
- height_power_comp_no_weighted_parameters_c2.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric power-law relationship linking tree height to tree diameter at breast height (equation 2) and with BAL as competition index (equation 4) and a no-weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2, comp corresponds to parameter cj defined in equation 4
- height_power_comp_RMSE_AIC_results.csv: mean ( _mean_) and total ( _total_) RMSE values, and mean and total AIC values, obtained over all 100 resamplings for each studied species with either a weighted or no weighted protocol effect included in the power-law relationship (equation 2), and with either BAT (c1) or BAL (c2) included as the competition index (see equation 4). Note that in the case of AIC, only the weighted protocol effect in the models was considered based on the results obtained with RMSE.
- No competition
- treeheight_mean_parameters_nocompetition.csv: average parameters obtained from 100 resamplings for each species studied and each of the parameters defined in the selected allometric asymptotic relationship linking tree height to tree diameter at breast height (i.e., weighted protocol effect); a1, a2 and a3 correspond respectively to parameters a1, a2 and 3 defined in the manuscript with equation 1; h stands for tree height and mean and sd represent the mean of the parameter and the standard deviation obtained over the 100 resamplings, respectively
- height_asympt_nocomp_no_weighted_parameters.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric asymptotic relationship linking tree height to tree diameter at breast height (equation 1) with a no-weighted protocol effect; a1, a2 and a3 correspond respectively to parameters a1, a2 and a3 defined in the manuscript with equation 1
- height_asympt_nocomp_weighted_parameters.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric asymptotic relationship linking tree height to tree diameter at breast height (equation 2) with a weighted protocol effect; a1, a2 and a3 correspond respectively to parameters a1, a2 and a3 defined in the manuscript with equation 1
- height_asympt_nocomp_RMSE_results.csv: mean ( _mean_) and total ( _total_) RMSE values obtained over all 100 resamplings for each studied species with either a weighted or no weighted protocol effect included in the asymptotic relationship (equation 1).
- height_power_nocomp_no_weighted_parameters.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric power-law relationship linking tree height to tree diameter at breast height (equation 2) with a no-weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2
- height_power_nocomp_weighted_parameters.csv: parameters obtained over the 100 resamplings fitted for each species studied and each of the parameters defined in the allometric power-law relationship linking tree height to tree diameter at breast height (equation 2) with a weighted protocol effect; a1 and a2 correspond respectively to parameters a1 and a2 defined in the manuscript with equation 2
- height_power_nocomp_RMSE_results.csv: mean ( _mean_) and total ( _total_) RMSE values obtained over all 100 resamplings for each studied species with either a weighted or no weighted protocol effect included in the power-law relationship (equation 2).
Light interception
- out_light_alone.csv: output file of the light interception experiment on an isolated tree with a crown representative of weak or strong local competition. An isolated tree of diameter ‘dbh_cm’ (15cm or 30cm) was placed in the middle of a 30m*30m (0.09ha) stand of species. For each tree, crown dimensions were calculated using the specific allometry for each of the 105 parameter combinations (‘id_allom’) to calculate the tree height (‘height_m’), crown radius (‘cradius_m’) and crown ratio (‘cratio_m’). Crown dimensions for an isolated tree were calculated theoretically undergoing competition associated with a BAL of either 0m2/ha (control, no competition) or 30m2/ha (‘bal_m2ha’). The shape of the crown was set according to the order of the species (‘order’: G = gymnosperm and A = angiosperm) with gymnosperms having a paraboloid crown shape (maximum diameter at the height of the base of the crown) and angiosperms having an ellipsoid crown shape (maximum crown diameter at half the depth of the crown). The ‘crown_lad’ variable (leaf area density) was set for all species at 0.5m2/m3. For each case, the SamsaraLight model was ran to calculate the light interception by the isolated tree over a full year at a given location. The total energy intercepted by the isolated tree ‘e’ (in MJ) and the potential energy intercepted by the tree in the absence of neighbours ‘epot’ (in MJ, here identical to ‘e’ because the tree is already alone in the stand) were recorded as ouptuts.
- out_light_compet.csv: light interception output file for a co-dominant or dominant tree in a theoretical stand with a basal area of varying size. Virtual stands of 30m*30m (0.09ha) were created with 6 modalities of mean square diameter Dg (15cm or 30cm) and total basal area BA in m2/ha (low, medium or high), leading to information indicated in the ‘id_pattern’ column. 20 stand structures, called patterns, are therefore simulated, with 10 repetitions (‘id_rep’) for each of the two Dg values. For each tree in each pattern, the associated BAL ‘bal_m2ha’ (basal area of the largest trees in the entire 1ha virtual stand) was calculated. Then, for each species and each pattern, its crown dimensions ‘h_m’, ‘cradius_m’ and ‘cratio’ were computed for each of the 105 allometry combinations ‘id_allom’ at a dbh equal to the Dg of the stand and a BAL competition equal to that experienced in the pattern in question. The stand patterns enabled to calculate the light interception of a competing tree for each species, with a fixed structure, since only the parameters of the allometry equations and the species order (which defines the allometry of the competitors) will have an influence on the light output variables. The LAD, the shape of the crowns and the conditions of the SamsaraLight model are identical to the previous experiment (out_light_alone.csv). The total energy intercepted by the isolated tree ‘e’ (in MJ) and the potential energy intercepted by the tree in the absence of neighbours ‘epot’ were recorded as ouptuts.
Traits and Functional groups
- functional_groups.csv: information on the taxonomic group (i.e. either angiosperms or gymnosperms) to which the species studied in this manuscript belong. Species names are specified in ‘sp’, and it taxonomic group in ‘group’ as A or G, as well as in ‘new_gp’ as angio or gymno. Both abbreviations correspond to the angiosperm and gymnosperm taxonomic groups, respectively.
- sp_traits_complete.csv: information on either the climate niche or functional traits of the studied species. Hence, species names are specified in ‘species’. Then, mean annual temperatures ‘mat’, mean annual precipitations ‘map’ and evapotranspiration ‘etp’ experienced by the species in their climatic niches are provided; along with information regarding key functional traits, i.e., specific leaf area ‘sla’, wood density ‘ssd’ and shade tolerance ‘shade_tol_mean’
Code/software
This dataset is based on csv files. The figures of the analysis can be replicated with R code hosted at https://github.com/LauraTouzot/Tree_crown_allometry_DECLIC_Project
Access information
Allometric equations were derived from these raw data:
Databases, repositories and data papers' references:
- BAAD: a Biomass and Allometry Database for woody plants (https://github.com/dfalster/baad)
- FHM: Forest Health Monitoring (https://www.fs.usda.gov/foresthealth/protecting-forest/forest-health-monitoring/index.shtml)
- FIA: Forest Inventory and Analysis (https://www.fs.usda.gov/research/inventory/FIA)
- Spanish National Forest Inventory (https://www.miteco.gob.es/en/biodiversidad/servicios/banco-datos-naturaleza/informacion-disponible/ifn3_bbdd_descargas.htm.aspx; https://www.miteco.gob.es/en/biodiversidad/servicios/banco-datos-naturaleza/informacion-disponible/ifn2_descargas.aspx)
- FunDivEUROPE: Functional significance of forest biodiversity (Germany: https://bwi.info/Download/de/BWI-Basisdaten/ACCESS2003/; Finland and Sweden:
https://doi.org/10.5061/dryad.wm37p⟨200b⟩vmkw)
- MONTANE (https://doi.org/10.57745/GFITOT)
- GenTree (Opgenoorth, L., Dauphin, B., Benavides, R., Heer, K., Alizoti, P., Martínez-Sancho, E., ... & Cavers, S. (2021). The GenTree Platform: growth traits and tree-level environmental data in 12 European forest tree species. GigaScience)
- LegacyTreeData (https://legacytreedata.org/)
- ICP Forests (International Co-operative Programme on Air Pollution Effects on Forests) (https://icp-forests.org/documentation/Introduction/index.html)
- EuMIXFOR (https://doi.org/10.5061/dryad.8v04m)
- French National Forest Inventory (https://inventaire-forestier.ign.fr/dataifn/)
- Canada’s National Forest Inventory (https://nfi.nfis.org/en/datarequest)
- Quebec’s National Forest Inventory (inventaires.forestiers@mffp.gouv.qc.ca)
- Tallo (https://zenodo.org/record/6637599#.ZFoYT-xBw-Q)
- Sample tree biomass data for Eurasian forests (https://elar.usfeu.ru/handle/123456789/4931)
- Harvard Forest Data Archive (https://harvardforest1.fas.harvard.edu/exist/apps/datasets/showData.html?id=HF339)
- Data papers
- Fuhr, M., Cordonnier, T., Courbaud, B., Kunstler, G., Mermin, E., Riond, C., & Tardif, P. (2017). Long‐term tree inventory data from mountain forest plots in France.
- Evans, M. R., Moustakas, A., Carey, G., Malhi, Y., Butt, N., Benham, S., Pallett, D. & Schäfer, S. (2015). Allometry and growth of eight tree taxa in United Kingdom woodlands. Scientific Data.
- Dettmann, G. T., & MacFarlane, D. W. (2019). Trans‐species predictors of tree leaf mass. Ecological Applications.
- Anderson‐Teixeira, K. J., McGarvey, J. C., Muller‐Landau, H. C., Park, J. Y., Gonzalez‐Akre, E. B., Herrmann, V., ... & McShea, W. J. (2015). Size‐related scaling of tree form and function in a mixed‐age forest. Functional Ecology.
- Dalponte, M., & Coomes, D. A. (2016). Tree‐centric mapping of forest carbon density from airborne laser scanning and hyperspectral data. Methods in Ecology and Evolution.
