Search for a command to run...
Data and code for the article titled "Small habitat patches can be the largest contributors to urban biodiversity across taxonomic groups" by Kilian Perrelet, Lauren M. Cook, Florian Altermatt, Federico Riva, and Marco Moretti. The information contained in this repository is referenced in the manuscript. ***************************************************** data/ The data/ folder contains processed data used to generate all figures (except Figure 1, which is a drawing). All files can be generated using dedicated scripts and raw data. patches.dbf - shapefile for all delinated patches. patches.prj - shapefile for all delinated patches. patches.shp - shapefile for all delinated patches. patches.shx - shapefile for all delinated patches. These files can be recreated using the 1_define_patches.R script. covariates.csv - covariate values extracted for all patches (table format). covariates.gpkg - covariate values extracted for all patches (geopackage format). Rows = sites, columns = covaiates, cell value = covariate value. These files can be recreated using the 2_calculate_covariates.R script. invertebrates.csv - species-by-site matrix for invertebrate taxa.vertebrates.csv - species-by-site matrix for vertebrate taxa.trees.csv - species-by-site matrix for all tree taxa.trees_native.csv - species-by-site matrix for native tree taxa.trees_nonnative.csv - species-by-site matrix for non-native tree taxa. Rows = sites, columns = species, 1 = presence, 0 = absence. These files can be recreated using the 3_prepare_df.R script. invertebrates_varsel.RData - results of variable selection for invertebrates. vertebrates_varsel.RData - results of variable selection for vertebrates. trees_varsel.RData - results of variable selection for trees. These files can be recreated using the 4_fit_gdm.R script. invertebrates_uncertainty.csv - results of I-spline bootstrapping for invertebrates. vertebrates_uncertainty.csv - results of I-spline bootstrapping for vertebrates. trees_uncertainty.csv - results of I-spline bootstrapping for trees. These files can be recreated using the 4_fit_gdm.R script. ***************************************************** raw/ The raw/ folder contains raw data used to generate files in the data/ folder. All datasets are publicly available at can be downloaded at: Habitat map: https://www.stadt-zuerich.ch/geodaten/download/Biotoptypenkartierung_2020?format=10007 Fauna data: https://data.stadt-zuerich.ch/dataset/geo_fauna_ogd Tree data: https://www.stadt-zuerich.ch/geodaten/download/bauminventar LiDAR data: https://www.swisstopo.admin.ch/en/lidar-data-swisstopo Invasive species: https://www.infospecies.ch/fr/neobiota/informations-especes-donnees-distribution.html Habitat classification: https://wwww.doi.org/10.1016/j.ecoinf.2024.102561 HabitatMap.cpg - polygon-based land-use map. HabitatMap.dbf - polygon-based land-use map. HabitatMap.prj - polygon-based land-use map. HabitatMap.shp - polygon-based land-use map. HabitatMap.shx - polygon-based land-use map. Rows = polygons. fauna_geodata.RData - shapefile for all observed faunal species. Rows = observations. trees_geodata.cpg - shapefile for all observed trees. trees_geodata.dbf - shapefile for all observed trees. trees_geodata.prj - shapefile for all observed trees. trees_geodata.shp - shapefile for all observed trees. trees_geodata.shx - shapefile for all observed trees. Rows = observations. infoflora_checklist.xlsx - list of non-native species for trees. invasive_species_list.csv - list of invasive species (all taxa). Unused. Rows = species. lidar.rds - lidar catalogue for the city of Zurich, preloaded in R. HabitatMapClass_Overview.csv - classification of green spaces into high- and low-maintained. Based on Dietzel, A., Moretti, M., and Cook, L. M. (2024). Shrinkage-based Bayesian variable selection for species distribution modelling in complex environments: An application to urban biodiversity. Ecological Informatics, 81, 102561. https://doi.org/10.1016/j.ecoinf.2024.102561 ***************************************************** fig2.R - generates Figure 2. fig3.R - generates Figure 3. fig4.R - generates Figure 4. figS1.R - generates Figure S1. figS2.R - generates Figure S2. figS3.R - generates Figure S3. figS4.R - generates Figure S4. figS5.R - generates Figure S5. figS6.R - generates Figure S6. figS7.R - generates Figure S7. figS8.R - generates Figure S8. figS9.R - generates Figure S9. figS10.R - generates Figure S10. figS11.R - generates Figure S11. figS12.R - generates Figure S12. figS13.R - generates Figure S13. figS14.R - generates Figure S14. figS15.R - generates Figure S15. figS16.R - generates Figure S16. 1_define_patches.R - delinates all patches. 2_calculate_covariates.R - calculates environmental conditions across all delinated patches. 3_prepare_df.R - transforms the observation data into species-per-site matrices. 4_fit_gdm.R - uses the species-per-site and covariate matrices to fit generalized dissimilarity modeling. ***************************************************** Data dictionary: FC_Grey | Fraction covered by impervious surfaces within each patch FC_Water | Fraction covered by water within each patch FC_Green | Fraction covered by any sort of green spaces within each patch FC_Forest | Fraction covered by forests within each patch FC_Green_Low | Fraction covered by low-maintained green spaces within each patch Grey_500 | Fraction covered by impervious surfaces (500 m buffer from patch margins) Water_500 | Fraction covered by water (500 m buffer from patch margins) Green_500 | Fraction covered by any sort of green spaces (500 m buffer from patch margins) Forest_500 | Fraction covered by forests (500 m buffer from patch margins) Green_Low_500 | Fraction covered by low-maintained green spaces (500 m buffer from patch margins) Area | patch area enn_mn | connectivity metric: mean nearest neighbor distance (500 m buffer from patch margins) ed | connectivity metric: edge density (500 m buffer from patch margins) cohesion | connectivity metric: patch cohesion (500 m buffer from patch margins) mesh | connectivity metric: effective mesh size (500 m buffer from patch margins) d2_Water | geographic distance to the nearest water body d2_Forest | geographic distance to the nearest forested habitat Total_meanH | mean vegetation height within each patch Total_sdH | standard deviation of the vegetation height within each patch Total_density | vegetation density within each patch PatchID | patch id