The data are derived from the PREDICTS Project database (https://data.nhm.ac.uk/dataset/902f084d-ce3f-429f-a6a5-23162c73fdf7) used within the paper titled "Human-dominated land uses favour species affiliated with more extreme climates, especially in the tropics" (DOI: 10.1111/ecog.04806). If using this data, please make sure that the PREDICTS Project is also cited.

For Williams_et_al_Ecography_data_CWM -> These data include, for the species assemblages in the analyses, the raw community weighted means (CWMs) for the extreme (maximum or minimum) and range-wide variation (standard deviations) for each climatic niche property. The columns for the CWMs of the extremes are headed CWM_MAXTmax, CWM_MINTmin, CWM_MAXPpmax and CWM_MINPpmin to match the terms used in the paper. The columns for the CWMs of range-wide variation are headed CWM_STDTmax, CWM_STDTmin, CWM_STDPpmax and CWM_STDPpmin. The columns headed 'Total_abundance' and 'Species_richness' were calculated during data preparation. The other columns included are details about each site, unchanged from the PREDICTS Project database.

For Williams_et_al_Ecography_data_Abundance_Tmax_Tropics, Williams_et_al_Ecography_data_Abundance_Tmax_Temperate, Williams_et_al_Ecography_data_Abundance_Ppmin_Tropics and Williams_et_al_Ecography_data_Abundance_Ppmin_Temperate -> These are the data used for the abundance analyses. There are four files within each of these files, one for each species group (formed by splitting around the medians of the exteme and standard deviation of the climatic variable in question, i.e. either maximum temperature of the hottest month or precipitation of the driest month). 'Tropics' or 'Temperate' within the file name denote whether the data are from from tropical or temperate latitudes, respectively. The column headed 'LogAbund' includes the log(x+1) transformation of the abundance measures, used within the models. The columns headed 'Simpson_diversity', 'ChaoR', 'Richness_rarefied', 'UseIntensity' and 'UI' are other variables we calculated along the way, but did not use in these analyses. The column 'LandUse' contains our slight alterations (e.g. changing vegetation to Vegetation) of the 'Predominant_land_use' column found within the PREDICTS Project database. The columns headed 'Total_abundance' and 'Species_richness' were calculated by us using the PREDICTS Project database. The other columns include unchanged species-level data from the PREDICTS Project database.