The global distribution of grass functional traits within grassy biomes

Aim The sorting of functional traits along environmental gradients is an important driver of community and landscape scale patterns of functional diversity. However, the significance of environmental factors in driving functional gradients within biomes and across continents remains poorly understood. Here, we evaluate the relationship of soil nutrients and climate to leaf traits in grasses (Poaceae) that are hypothesized to reflect different strategies of resource use along gradients of resource availability. Location Global. Taxon Poaceae. Methods We made direct measurements on herbarium specimens to compile a global dataset of functional traits and realized environmental niche for 279 grass species that are common in grassland and savanna biomes. We examined the strength and direction of correlations between pairwise trait combinations and measured the distribution of traits in relation to gradients of soil properties and climate, while accounting for phylogenetic relatedness. Results Leaf trait variation among species follows two orthogonal axes. One axis represents leaf size and plant height, and we showed positive scaling relationships between these size‐related traits. The other axis corresponds to economic traits associated with resource acquisition and allocation, including leaf tensile strength (LTS), specific leaf area (SLA) and leaf nitrogen content (LNC). Global‐scale variation in LNC was primarily correlated with soil nutrients, while LTS, SLA and size‐related traits showed weak relationships to environment. However, most of the trait variation occurred within different vegetation types, independent of large‐scale environmental gradients. Main conclusions Our work provides evidence among grasses for relationships at the global scale between leaf economic traits and soil fertility, and for an influence of aridity on traits related to plant size. However, large unexplained variance and strong phylogenetic signal in the model residuals imply that at this scale the evolution of functional traits is driven by factors beyond contemporary environmental or climatic conditions.