Abstract :
[en] Plants are able to synthesise and release volatile organic compounds (VOCs) aboveground (leaves, stems, flowers and fruits) and belowground (roots). Once emitted, these molecules are key mediators in biotic interactions as they can be perceived by plant neighbours (first trophic level) and are able to attract/repel organisms of the second (insect herbivores, plant parasitic nematodes) or the third trophic level (entomopathogenic nematodes, parasitoids, etc.). Although many laboratory and field experiments have focused on VOC-mediated plant-plant interactions aboveground, less is known regarding the roles played by root-emitted VOCs in between- and within-plant signalling. In this context, the main goals of this PhD thesis were to (1) identify and quantify the VOCs emitted by barley and chamomile roots and (2) study the influence of chamomile root volatiles on the growth (biomass production and allocation) and the root system architecture (RSA) of barley (interspecific model). Root-emitted VOCs were analysed without extracting the roots from the soil (in situ) using a three-step gas chromatography-mass spectrometry methodology. Plant-plant interaction bioassays were performed using an original experimental device allowing the controlled exposition of growing barley roots to the volatile compounds emitted by chamomile roots for 15 days. In order to speed up the RSA analysis of recipient barley plants, we developed an R package (archiDART) allowing (1) the batch processing of the raw data exported by Data Analysis of Root Tracings (DART) and root image analysis software tools supporting the Root System Markup Language (RSML) format, and (2) the automated computation of RSA traits.
Our results showed that crushed barley roots produced mainly hexanal, (E)-hex-2-enal, (E)-non-2-enal and (E,Z)-nona-2,6-dienal. Three-day-old seminal roots were characterised by higher total and individual VOC concentrations compared with older phenological stages. Our experiments also showed that enzymatic activities were required for volatile production. For each developmental stage, the lipoxygenase (LOX) specificity was greater for linoleic acid than for α-linolenic acid. The greatest LOX activities using linoleic and α-linolenic acids as substrates were measured in 7- and 3-day-old roots, respectively. Although undamaged barley roots did not release detectable amounts of VOCs, the analysis of VOCs emitted by mechanically injured roots showed that (E)-non-2-enal (13.8 ± 4.9 ng/g dry wt/h) and (E,Z)-nona-2,6-dienal (4.7 ± 1.8 ng/g dry wt/h) were the only VOCs detected in the plant rhizosphere. Contrasting with these results, the undamaged roots of 61- to 78-day-old chamomile plantlets released mainly one trinorsesquiterpene (albene) and four tricyclic sesquiterpene hydrocarbons (silphinene, modheph-2-ene, α-isocomene and β-isocomene) associated with the Asteraceae family. For each sesquiterpene hydrocarbon, the emission rate was positively correlated with plant age.
Based on these results, we performed plant-plant interaction bioassays to investigate the roles played by chamomile root volatiles on the growth and RSA of barley. After 15 days of exposure, plants exposed to the volatiles emitted by the soil and chamomile roots or by the soil alone (control) were morphologically similar. Although not statistically significant (P < 0.09), the leaf area and the total seminal root length were the only parameters that tended to be greater in plants that received the volatile compounds emitted by chamomile roots compared with control plantlets.
All these results are discussed in the context of belowground chemical ecology. In addition, some improvements of the experimental devices developed in this research project are also suggested at the end of this PhD thesis.
