Background Aboveground, plants release volatile organic compounds (VOCs) that act as chemical signals between neighbouring plants. It is now well documented that VOCs emitted by the roots in the plant rhizosphere also play important ecological roles in the soil ecosystem, notably in plant defence because they are involved in interactions between plants, phytophagous pests and organisms of the third trophic level. The roles played by root-emitted VOCs in between-and withinplant signalling, however, are still poorly documented in the scientific literature. Scope Given that (1) plants release volatile cues mediating plant-plant interactions aboveground, (2) roots can detect the chemical signals originating from their neighbours, and (3) roots release VOCs involved in biotic interactions belowground, the aim of this paper is to discuss the roles of VOCs in between-and withinplant signalling belowground. We also highlight the technical challenges associated with the analysis of root-emitted VOCs and the design of experiments targeting volatile-mediated root-root interactions. Conclusions We conclude that root-root interactions mediated by volatile cues deserve more research attention and that both the analytical tools and methods developed to study the ecological roles played by VOCs in interplant signalling aboveground can be adapted to focus on the roles played by root-emitted VOCs in between-and within-plant signalling.

1. Plant species that arrive first in the system can affect assembly (priority effects).However, effects of order of arrival of different plant functional groups (PFGs) on root development have not yet been investigated under field conditions. 2. We measured standing and fine root length density in the first and third year of a grassland field experiment. We wanted to know if manipulating PFG order of arrival would affect root development, and if priority effects are modulated by soil type.3. Sowing legumes first created a priority effect that was found in the first and third year, with a lower standing root length density in this treatment, even though the plant community composition was different in each of the studied years. Fine root length density was not affected by order of arrival, but changed according to the soil type. Synthesis.We found strong evidence that sowing legumes first created a priority effect below-ground that was found in the first and third year of this field experiment, even though the functional group dominance was different in each of the studied years. K E Y W O R D Sbelow-ground productivity, facilitation, historical contingency, order of arrival, plant functional groups, priority effects

Diversity of species and order of arrival can have strong effects on ecosystem functioning and community composition, but these two have rarely been explicitly combined in experimental setups. We measured the effects of both species diversity and order of arrival on ecosystem function and community composition in a grassland field experiment, thus combining biodiversity and assembly approaches. We studied the effect of order of arrival of three plant functional groups (PFGs: grasses, legumes, and non-leguminous forbs) and of sowing low and high diversity seed mixtures (9 or 21 species) on species composition and aboveground biomass. The experiment was set up in two different soil types. Differences in PFG order of arrival affected the biomass, the number of species and community composition. As expected, we found higher aboveground biomass when sowing legumes before the other PFGs, but this effect was not continuous over time. We did not find a positive effect of sown diversity on aboveground biomass (even if it influenced species richness as expected). No interaction were found between the two studied factors. We found that sowing legumes first may be a good method for increasing productivity whilst maintaining diversity of central European grasslands, although the potential for long-lasting effects needs further study. In addition, the mechanisms behind the non-continuous priority effects we found need to be further researched, taking weather and plant-soil feedbacks into account.