Trophic relationships between host and parasitic plants : a case study with the parasitic plant species <em>Phelipanche ramosa</em>

SPE Session S3B EA EcolDur Agrosup CT1National audiencePhelipanche ramosa (L.) Pomel is the most frequent and harmful parasitic plant species in France. It reproduces not only on crop species but also on many families of weeds. Thus, even in the absence of a host crop, weeds can allow the increase of the P. ramosa soil seed bank, which may then infest a subsequent crop. The holoparasitic plant P. ramosa attaches to the roots of its host and extracts all its nutrients from the host. Consequently, host/parasite trophic relationships are crucial in determining the growth and seed production of the two plants interacting. The objective of the present study was to quantify the intensity with which P. ramosa draws carbon assimilates from its host and to analyze whether this intensity varies with (1) host species, (2) the phenological stage of the host plant, and (3) plant growth conditions. A greenhouse experiment was conducted using three host species: oilseed rape (Brassica napus L.), the preferred P. ramosa host crop species and two host weed species, namely Capsella bursa-pastoris (L.) Medik. and Geranium dissectum L.. Plants of the three species were grown with or without P. ramosa seeds. They were grown under two contrasted light levels (100% and approximately 30 % daylight) in order to evaluate the robustness of our results under contrasting growth conditions. Plant biomass was considered as a relevant trait to analyze carbon fluxes between the two plants interacting as P. ramosa extracts all its nutrients from its host. So, host and parasite biomasses were measured separately at four phenological stages of the host species, from rosette to fructification stage. Host biomass was analyzed in the presence and absence of P. ramosa. The biomass of P. ramosa was analyzed as a function of the biomass of its host. Whatever the host species and the light level, the presence of P. ramosa significantly reduced the biomass of its host, with the total host+parasite biomass at a given date being lower than the biomass of the parasite-free host. The percentage of host biomass reduction due to the parasite showed that the negative impact of P. ramosa on host biomass increased during host growth and that the three host species differed in sensitivity to P. ramosa. Whatever the light level, C. bursapastoris and oilseed rape were, respectively, the least and the most sensitive host species. Whatever the phenological stage and the light level, the biomass of P. ramosa was the highest on oilseed rape, and this both in absolute value and relatively to host biomass. From our results, equations were proposed to predict both parasite biomass and host biomass reduction due to parasite as a function of host species identity, phenological stage and light level. In the near future, results will be synthesized into a mechanistic model of the effect of cropping systems on pest dynamics. This model will help to design new strategies for the management of P. ramosa by identifying which techniques must be chosen and combined to optimize the control of both the parasite and susceptible weeds