A fully mapped community approach to unveil the effects of biotic interactions

At small scales where biotic interactions are strongest, constraints on coexistence can drive similarity or dissimilarity in functional traits, according to the theories of weaker competitor exclusion, and limiting similarity, respectively. Neighbourhood density can also be correlated with functional traits, reflecting the demands of a more competitive environment. While previous studies in grassland communities have found evidence of limiting similarity, little attention has been given to scales under 20 cm, where interactions could be strongest. To fill those gaps we treated the community as a marked point pattern, by fully mapping the position of each functional individual in two semi-natural dry grasslands (“open” and “closed”) using 10 50 × 50 cm quadrats. We collected trait information for 1094 functional individuals, namely height, leaf area and specific leaf area, covering independent axes of variation in plant size and resource economics. Having individual-level trait measurements allows for estimation of intraspecific trait variability (ITV), which is often neglected and potentially high, and to conduct a purely functional analysis, without relying on species-level mean trait information. Overlap-based dissimilarity between species and grassland type confirmed the large contribution of ITV. Many point pattern functions didn’t deviate from the null model, but in the closed grassland, significant trait similarity and negative density correlation was found for traits related to plant size; the multivariable approach considering distances in the trait space showed trait similarity in the open grassland. The point pattern approach proves to bring new insights into assembly rules and scales of interest for grasslands.