Trait variation explained by climbing mechanisms
We found functional traits for 755 liana species: 551 representing active climbing species and 204 representing passive climbing species. Species employing an active climbing mechanism were characterized by traits associated with acquisitive strategy compared to species with a passive climbing mechanism (Fig. 2). Active climbing species showed higher specific leaf area (t-test, P = 0.0001; active = 187, passive = 57), larger total leaf area (t-test, P = 1.7e-18; active = 380, passive = 136), higher leaf nitrogen content (t-test,P = 0.02; active = 129, passive = 38) and slightly higher maximum photosynthetic rate per mass unit (t-test, P = 0.05; active = 56, passive = 20) compared to passive climbing species (Fig. 2a-d). We did not find differences in wood density between the active and passive climbing species (t-test, P = 0.2; active = 119, passive = 19) (Fig. 2e). Seed mass was higher in the active compared to the passive climbing species (t-test, P = 0.0006; active = 225, passive = 69) (Fig. 2f). Overall, only the passive climbing species showed a significant phylogenetic signal (Blomberg“s K ) considering the functional traits we analyzed here (Appendix S2, Table S2). Passive climbing species showed significant phylogenetic signals for SLA, LA, Nmass and seed mass. These results indicate that active climbing species tended to show patterns of trait variation among species that were independent of phylogenetic relatedness.