Intransitivity increases plant functional diversity by limiting dominance in drylands worldwide

Biotic interactions are key determinants of plant community structure. Indirect interactions such as intransitivity (i.e., the absence of competitive hierarchies among species) have been hypothesized to benefit diversity within plant communities. However, their effect on functional diversity remains scarcely explored in real communities. Here, we develop a novel approach to infer intransitivity from plant spatial patterns and functional traits (height and specific leaf area) and quantify its effect on different components of plant diversity along environmental gradients in 100 drylands from all continents except Antarctica. We first calculated the spatial association pattern for all perennials to infer competition between species. Trait values were used as a proxy of competitive hierarchies to infer the direction of these interactions. We used multiple regression models to evaluate how intransitivity responds to environmental variables (mean annual temperature and precipitation, precipitation seasonality, soil pH, sand content and woody cover). We also used confirmatory path analysis to evaluate the effects of intransitivity on species richness and evenness, trait dispersion and functional diversity. Intransitivity mostly responded to climatic variables and significantly increased with precipitation scarcity and seasonality. We found that intransitivity had significant effects on functional diversity, mostly by promoting plant community evenness. However, the dominance of woody vegetation (steppes versus shrublands) modulated this effect. Synthesis. Intransitivity increased the functional diversity of drylands, particularly under high rainfall seasonality, by limiting functionally dominant species. Our findings specify how intransitivity structures the functional diversity of dryland vegetation worldwide. Intransitivity may be particularly important in ecosystems where the availability of abiotic resources changes over time, thereby breaking down inherent competitive hierarchies between plant species. Neglecting intransitivity will bias our estimation of the impacts of biotic interactions on plant communities, a fundamental issue to fully understand how plant communities will respond to ongoing environmental changes.