Trait-by-species data by dataset

The sheets following present the trait data used for each dataset (labelled as the first author of the dataset's published source, separated by taxonomic group for the study that sampled multiple groups). Here, a row is a species and a column is a trait. The names of each trait are preceded by a four-letter code that represents the trait class within which the trait is classified (and where it was weighted accordingly in analyses). These classes (and their abbreviations) are: body dimensions (body), dispersal (disp), diet and feeding (diet), habitat and environmental preferences (habt), phenology (phen) and reproduction (repr). For macroinvertebrate datasets (the next 10 sheets), 'taxa' refers to the organism as identified in the original published study, 'database_entry' refers to the trait data entry to which the organism was matched. In all studies, 'estimation_level' is the taxonomic level from which trait data were sourced (and estimated, in cases where data were sourced at a taxonomic resolution higher than that reported in the original study). Common names are included for some studies. More detailed descriptions of traits are listed in Appendix S4 in Supporting Information, and details of trait data collection are explained in 'Data collection for traits' in the Methods section of the full text.