Understanding the determinants of species distributions is a central topic in ecology. Competition, stress tolerance and colonization, respectively represented by Grime’s competitor (C), stress-tolerator (S) and ruderal (R) schemes, are three important functions that interactively influence plant distributions. In this study, we compiled a dataset of 2645 vascular plant species to explore the roles of the CSR strategies in global plant distribution. We analyzed the associations between the CSR scores and species range size with phylogenetic generalized least square (PGLS) models and phylogenetic path analysis, both of which accounted for the effects of species phylogenetic relatedness, longevity, and growth form. The functional strategy-range size associations differed across different distributional ranges and growth forms. Specifically, species global and native range sizes were positively associated with the R score; species naturalized range size was positively associated with the C score; and all range-size measurements were negatively associated with the S score. These patterns were mostly driven by herbs but not shrubs or trees. For species global and native-range distributions, the patterns of shrubs were even opposite to those of herbs. Our work emphasizes the importance of distinguishing the functional strategy-distribution associations between different distributional ranges and growth forms for ecosystem conservation and invasion risk prediction, because of the trade-offs among the CSR strategies.

All range sizes were estimated based on equal-area projection of species occurrence records from the Global Biodiversity Information Facility (GBIF). We first downloaded GBIF records using he function  'occ_search' in the R package 'rgbif'. Then, we projected the records based on World Eckert IV system and assign the records to species native or naturlized range by referring to the Global Naturalized Alien Flora (GloNAF) database. Over-estimation of naturalized range size will occur, when we designated a whole continent as the naturalized range, while, in fact, the focal species is native to some but non-native to the other regions in the corresponding continent. Therefore, we also supplemented evaluation of the the risk level of such over-estimation by calculating the geographic distances between the regions in the GBIF that have not been included in the GloNAF and those included in the GloNAF. This was done using the 'distGeo' function from the 'geosphere' package.

For the CSR scores, we directly acquired the data from the global dataset compiled by Pierce et al. (2017), which included the C, S and R scores for 3068 vascular plant species. Data on species longevity and growth forms were also taken from Table S1 of Pierce et al. (2017) after making some modifications on the longevity and growth forms. We designated annuals or biennials as short-lived species and perennials as long-lived species. In addition, we combined forbs, graminoids, herbaceous vines and aquatic species as herbs, and combined lianas and shrubs as shrubs.