Les interactions souterraines entre plantes incluent des échanges de nutriments

[Departement_IRSTEA]Territoires [ADD1_IRSTEA]Dynamique et fonctionnement des écosystèmesInternational audienceTwo distinct mechanisms of plant - plant competition have been classically distinguished, competition due to exploitation, the fast pre-emption of available resources by a plant to a level that does not allow the neighbouring plant to grow, and competition due to interference, the direct control of a plant by its neighbour, for example through secretion of allelopathic compounds. However recent literature suggests that interactions between plants are much more complex and include exchange of information, molecule and resources, for example through the mechanism of rhizodeposition. We tested the hypothesis that the competition between oak seedlings (Quercus petraea) and a common perennial grass (Molinia caerulea) during forest natural regeneration does include a competition by exploitation of nitrogen (N) resources but also exchange of N between both species to the benefit of the grass. Both species were cultivated together in 20 L pots, either directly in contact or separated by a membrane with different pore sizes, so that roots of both species can cross the membrane, or only hyphae of mycorrhiza, or only molecules. Oak seedlings were on a drip with a 15N enriched solution of urea (cotton-wick technique) at the beginning of the growing season. Oak, grass, and soil were harvested 2, 4, 6, and 10 months after start of labelling. Samples were analysed for N content and 15N excess. The presence of grass reduced oak growth as expected. However the presence of oak favoured the grass growth as compared when its neighbour was a grass tuft (intraspecific competition). 15N was found in all oak organs (leaf, stem and root), but also in the soil and grass organs (tuft and roots), as soon as 2 months after the start of labelling. 15N was also retrieved in the grass after 10 months, though in a low quantity. 15N was found in the grass for every tested membrane pore size, meaning that 15N circulated in both sides of the membrane without any root contact, yet chemical forms of N has not been identified (organic compounds?). Those results suggest that beyond the exploitation of N by the grass to the detriment of oak seedlings, grass also benefited from a release of N by oak. The underlying mechanisms are not fully understood but they emphasize the important role of rhizodeposition as well as trophic players such as mycorrhizas, and microorganisms. They also suggest that exchange of nutrients, other molecules and information are much more common than suspected in belowground plant interactions