Data and code from: The overlooked link between different resource partitioning strategies and plant species richness in tropical alpine ecosystems
Data files
Dec 10, 2025 version files 7.04 MB
-
README.md
3.78 KB
-
Supplementary_Material__S3_associated_data.zip
7.04 MB
Abstract
Species co-existence is based on resource partitioning and modulates biodiversity patterns across climates, latitudes, and altitudes. Resource partitioning can occur via niche size or separation in the geographic range or ecological niche. While resource partitioning promotes biodiversity, the impact of different partitioning strategies on species richness remains largely unexplored. Study locations: two ecosystems with similar climates and ages, the species-rich tropical alpine ecosystem in the South American Andes and the more species-poor tropical alpine ecosystem in the eastern African mountains. Study time period: present-day distribution and climatic conditions, integrating phylogenetic information extending back to the last 7 million years maximum. Major taxa studied: Six lineages from the Asterales; three in each ecosystem, respectively. We test whether geographic range and climatic niche partitioning strategies may explain differences in species richness between the two ecosystems. We combine phylogenomic data with occurrence records and estimate metrics of size and overlap for climatic niche and geographic range. We show that the Andean species have larger climatic niches than the African species, suggesting that niche size is not explaining higher species richness in the Andes. Instead, a striking pattern for species with overlapping geographic ranges emerged: the Andean species show less climatic niche overlap than the African species, indicating more effective niche separation among Andean species. We hypothesise that a different pattern of resource partitioning, specifically increased niche separation among geographically overlapping species in the Andes compared to the eastern African mountains, contribute to the species richness difference between these tropical alpine biodiversity hotspots.
Data
We studied six Asterales lineages in the tropical alpine ecosystems of South America and East Africa, i.e., Loricaria, Senecio, and Oritrophium in South America and Dendrosenecio, Helichrysum, and Lobelia in East Africa. We compiled occurrence information from herbaria and supplemented it with GBIF data (https://doi.org/10.15468/dd.8cpqns). The table contains all columns as provided by GBIF (simple format; https://techdocs.gbif.org/en/data-use/download-formats#simple), for the occurrence data from herbaria, some column information was not available and is coded as NA (not available).
Phylogenetic data for each lineage are based on Compositae1061 loci (Mandel et al., 2014); newly sequenced material is available under Bioproject PRJNA1092049. For Dendrosenecio, Loricaria, and Oritrophium, existing phylogenetic estimates were available, and the respective authors shared them. For the remaining lineages, we reconstructed phylogenetic trees using RAxML-NG that were then made ultrametric using treePL. Each lineage has its own folder that contains the respective data in a computer readable formats.
Supplementary_Material__S3_associated_data.zip
|- Dendrosenecio/
| |- raxmlng.support.reroot
| |- treepl_Dendrosenecio.tre
|- Helichrysum/
| |- concatenated50_70.fasta
| |- RAxMLpartitions.txt
| |- raxmlng.support.reroot
| |- treepl_Helichrysum.tre
|- Lobelia/
| |- concatenated50_70.fasta
| |- RAxMLpartitions.txt
| |- raxmlng.support.reroot
| |- treepl_Lobelia.tre
|- Loricaria/
| |- concatenated50_70.fasta
| |- RAxMLpartitions.txt
| |- raxmlng.support.reroot
| |- treepl_Loricaria.tre
|- Oritroppl_Oritrophium.tre
|- Senecio/
| |- concatenated50_70.fasta
| |- RAxMLpartitions.txt
| |- raxmlng.support.reroot
| |- treepl_Senecio.tre
|- cleaned_occ_v1.0.csv
All niche and range analyses conducted for the associated publication are available via R scripts submitted to Zenodo (NicheSizeDifferences-main.zip). Scripts are numbered from zero to seven and reflect order of execution.
NicheSizeDifferences-main.zip
|- FUN/
| |- 00_all_myfun_combined.R
| |- 00_func_red_code.R
| |- 00_fun_for_global.R
| |- 00_load_packages_kubernetes.R
| |- 001_get_started.R
| |- 01a_clean_occ.R
| |- 01d_clean_occ_ingroup.R
| |- 03a_prepare_pcaallenv_occonly.R
| |- 03b_plot_pca.R
| |- 04a_calc_geo_area.R
| |- 04b_make_zscore.R
| |- 04c_calc_niche_area.R
| |- 04d_calc_SchoenersD.R
| |- 04e_pgls_looop.R
| |- 04f_combine_data.R
| |- 05_permutation_niche.R
| |- 05_permutation_range.R
| |- 06_gendist_niche.R
| |- 07_habitat_heterogeneity_rasterdiv.R
|- LICENSE
|- README.md
- Scripts starting with 00_* contain different functions that are needed.
- Script 001_get_started.R needs to be adapted to reflect working directory, path to climatic data and elevation as well as to phylogenetic trees.
- Scripts starting with 01* are used to clean occurrence information.
- Scripts starting with 03* are used to make a PCAenv.
- Scripts starting with 04* should be executed in order to calculate range size, niche size, niche overlap, PGLS analyses, and combine all results into a data-frame.
- Script starting with 05* were used to permute niche and range size.
- Script 06_gendist_niche.R calculates the genetic distances in relation to niche and range characters.
- Script 07_habitat_heterogeneity_rasterdiv.R calculates the different estimates of habitat heterogeneity.
Further details on how datasets were generated can be found in the associated article.
